Surface Water Quality Assessment and Modelling A case ... · PDF fileSurface Water Quality Assessment and Modelling A case study in the Tuul River, Ulaanbaatar city, Mongolia By Ochir

Surface Water Quality Assessment and Modelling A case study in the Tuul River, Ulaanbaatar city,

Mongolia

Ochir Altansukh

March 2008

Surface Water Quality Assessment and Modelling A case study in the Tuul River, Ulaanbaatar city,

Mongolia

By

Ochir Altansukh

Thesis submitted to the International Institute for Geo-information Science and Earth Observation in

partial fulfilment of the requirements for the degree of Master of Science in Geo-information Science

and Earth Observation, Specialisation: Integrated Watershed Modelling and Management

Thesis Assessment Board Professor. Dr. Z. Bob Su Chairman WREM Department, ITC

Associate Professor. Dr. Ir. C.M.M. Chris Mannaerts First Supervisor WREM Department, ITC

Assistant Professor. Dr. Ir. Mhd. Suhyb Salama Supervisor WREM Department, ITC

Assistant Professor. Dr. Ir. D.C.M. Denie Augustijn External Examiner WEM Department, UT

INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION

ENSCHEDE, THE NETHERLANDS

Disclaimer This document describes work undertaken as part of a programme of study at the International Institute for Geo-information Science and Earth Observation. All views and opinions expressed therein remain the sole responsibility of the author, and do not necessarily represent those of the institute.

To my father, Ochir’s family.

You are all special to me.

i

Acknowledgements

First of all, my deepest thanks to my father, Ochir’s family, including my mother, my sisters, my

brothers, my wife and my son, my younger sisters, my younger brothers for their honourable support,

encouragements and love.

Specially, all my infinite love, sincere gratitude and unconditional faith to my Mandaa for her pure

heart, unlimited support and eternal love.

Special thanks to Dr. J.L. John van Genderen, who is a professor at ITC, Department of Earth

Observation Science, for his vital support to “open the gate” and kind-heart to the Geo-science in

Mongolia.

I wish to express my honest gratitude to Ms. D.Narantuya, a leader of NGIC for NRM project, and to

Mr. D.Amarsaihan, a training coordinator of NRM project, for their support to my research and

extension of study.

I am grateful to Ir. A.M. Arno van Lieshout, a programme director of WREM, and Dr. H.A.M.J. Hein

van Gils, a professor at ITC, NRM, who gave me very nice opportunity to study at ITC in MSc

course.

Datasets used in this study have been obtained from different organizations and persons, including

NAMHEM, CLEM, Ms. Munhtsetseg (Tuul-UB hydrological station), Mr. G.Davaa (WS,

NAMHEM), Ms. Ya.Erdenebayar (CLEM) and Mr. J.Bayasgalan (NGIC project). We would like to

thank them all for making available data.

I do particular thank to professor Dr. Ir. C.M.M. Chris Mannaerts, my supervisor, for his time,

comments, valuable ideas, support and guidance throughout the research. I wish to thank Dr. Ir. Mhd.

Suhyb Salama, my supervisor, for his ideal comments and time to read my thesis.

I appreciate to all my country mates, Ms. B.Orgil, Ms. B.Oyundari, Ms. M.Munguntuya, Mr.

D.Bayarbaatar, Ms. B.Tuul, and my classmates and friends, above all Ms. Le Thi Hanh, Ms. Nguyen

Phuong Tam. I will never forget your support, friendship and very nice time that we were shared

together at ITC.

Thanks to all ITC staffs, especially to WREM staffs for their important support and very advanced

lectures, classes during my study.

On top of all, thank very much to my God.

iii

Abstract

This research was conducted to assess the stream water quality of the Tuul River around the

Ulaanbaatar city, Mongolia and to model the impacts of wastewater treatment plants on the Tuul

River water quality, hereby mainly focusing on the CWTP (Central Wastewater Treatment Plant).

The assessment of surface water quality for natural river waters, was done using a water quality index,

developed by Erdenebayar,Ya., and Bulgan,T and the Mongolian National Standard MNS 4586-98.

This national standard was enacted by the CSM, Centre of Standardization and Measurement in 1998,

and represents the standards for maximum permissible levels of chemical variables in the surface

waters in Mongolia.

Geo-statistical techniques were utilized to estimate the spatial and temporal variability of the surface

water quality index, which was calculated by combination of 6 parameters, i.e. ammonia, nitrate,

nitrite, dissolved oxygen, chemical oxygen demand and biochemical oxygen demand. The monitoring

period covered 11 years from 1996 to 2006. The time series of water quality maps for the Tuul River

were also visualized using ILWIS.

DMS (Duflow Modelling System) was then used as tool for modelling of the river water quality. The

DO (dissolved oxygen) in the Tuul River network system was selected as main state variable. The DO

was modelled, calibrated and validated using an oxygen quality model (WUR, 2002) permitting to

analyse oxygen depletion by nitrogen and carbonaceous waste inputs in the river. Water quality

simulations were performed using hydro-chemical, hydraulic, climatic datasets between 2005 and

2006. The calibrated water quality model was then applied to evaluate the impact scenarios (of the

CWTP of Ulaanbaatar), for improvement of river water quality in order to meet the Mongolian

National Standards.

Keywords;

river water quality assessment, chemical datasets, hydraulic data, climatic data, pollution map,

contamination source, flow model, quality model, model calibration, model validation, sensitivity

analysis

v

Table of contents

Acknowledgements .................................................................................................................................. i Abstract .................................................................................................................................................. iii List of tables.......................................................................................................................................... vii List of figures....................................................................................................................................... viii List of equations.......................................................................................................................................x List of appendices .................................................................................................................................. xi Abbreviations ........................................................................................................................................ xii Chapter 1: Introduction ............................................................................................................................1

1.1: Background information ...............................................................................................................1 1.1.1: Spatial information ................................................................................................................1 1.1.2: Hydraulic information ...........................................................................................................2 1.1.3: Climatic information..............................................................................................................4 1.1.4: Hydro-chemical information..................................................................................................5

1.2: Recent problem statement.............................................................................................................5 1.3: Reason of topic selection ..............................................................................................................6 1.4: Research object .............................................................................................................................7 1.5: Research targets ............................................................................................................................7 1.6: Previuos studies ............................................................................................................................7

1.6.1: Quality studies using chemical method .................................................................................8 1.6.2: Quality studies using biological method................................................................................8

1.7: Research objectives.......................................................................................................................9 1.8: Research questions......................................................................................................................10 1.9: Research phases ..........................................................................................................................10 1.10: Methodology .............................................................................................................................11 1.11: Research hypothesis..................................................................................................................12 1.12: Outline of the thesis ..................................................................................................................12

Chapter 2: Field survey and data collection...........................................................................................13 2.1: Field survey.................................................................................................................................13

2.1.1: Selection of sampling site....................................................................................................13 2.1.2: Sampling method .................................................................................................................15 2.1.3: Chemical analysis in field....................................................................................................15

2.2: Data collection ............................................................................................................................16 2.2.1: Chemical data ......................................................................................................................16 2.2.2: Hydraulic data......................................................................................................................17 2.2.3: Climatic data........................................................................................................................17 2.2.4: Map and standards ...............................................................................................................18 2.2.5: Pollution source ...................................................................................................................19

Chapter 3: Water quality assessment .....................................................................................................21 3.1: Method for surface water quality assessment .............................................................................21 3.2: Basic geo-statistical analysis ......................................................................................................22

3.2.1: Descriptive statistics ............................................................................................................22 3.2.2: Exploratory graphics............................................................................................................24

3.3: Spatial water quality assessment.................................................................................................25 3.4: Temporal water quality assessment ............................................................................................27

3.4.1: Seasonal water quality assessment ......................................................................................30 3.5: Chapter conclusions....................................................................................................................31 3.6: Limitation in WQA.....................................................................................................................32

vi

Chapter 4: River flow model ................................................................................................................. 33 4.1: General introduction of flow model........................................................................................... 33 4.2: Background theory of flow simulation....................................................................................... 33

4.2.1: Saint-Venant equation......................................................................................................... 33 4.2.2: Manning equation ............................................................................................................... 34 4.2.3: Froude number .................................................................................................................... 35 4.2.4: Nash-Sutcliffe coefficient ................................................................................................... 36 4.2.5: Statistical methods ME, MAE, RMSE ............................................................................... 37

4.3: Flow model setup ....................................................................................................................... 37 4.3.1: Network development ......................................................................................................... 38 4.3.2: Initial condition of flow model ........................................................................................... 40 4.3.3: Boundary condition of flow model ..................................................................................... 41 4.3.4: Calculation setting .............................................................................................................. 42

4.4: Calibration.................................................................................................................................. 42 4.4.1: Sensitivity analysis.............................................................................................................. 44

4.5: Validation................................................................................................................................... 45 4.6: Chapter conclusion..................................................................................................................... 46 4.7: Limitation in a flow modelling .................................................................................................. 46

Chapter 5: Water quality model ............................................................................................................ 47 5.1: Introduction of quality model..................................................................................................... 47 5.2: Background theory of quality modelling ................................................................................... 48

5.2.1: Mass transport equation ...................................................................................................... 48 5.2.2: Peclet number...................................................................................................................... 50 5.2.3: DO balance equations ......................................................................................................... 50

5.3: Quality model setup ................................................................................................................... 53 5.3.1: Quality description.............................................................................................................. 55 5.3.2: Initial condition of quality model ....................................................................................... 55 5.3.3: Boundary condition of quality model ................................................................................. 56 5.3.4: Parameters in quality model................................................................................................ 57 5.3.5: External variables................................................................................................................ 57 5.3.6: Calculation setting .............................................................................................................. 58

5.4: Calibration.................................................................................................................................. 58 5.4.1: Sensitivity analysis.............................................................................................................. 59

5.4.1.1: Very sensitive parameters ............................................................................................ 59 5.4.1.2: Less sensitive parameters............................................................................................. 60 5.4.1.3: Very less and non sensitive parameters ....................................................................... 61

5.5: Validation................................................................................................................................... 62 5.5.1: Scenarios for SWQ improvement ....................................................................................... 63

5.6: Chapter conclusion..................................................................................................................... 64 5.7: Limitation in quality modelling ................................................................................................. 64

Chapter 6: Conclusions and recommendations ..................................................................................... 65 6.1: Overall conclusions.................................................................................................................... 65 6.2: Recommendations ...................................................................................................................... 66 6.3: Future research........................................................................................................................... 67 6.4: Output significance .................................................................................................................... 67 6.5: Limitations of the study ............................................................................................................. 67

References ............................................................................................................................................. 69 Appendices ............................................................................................................................................ 73

vii

List of tables

Table 1: Annual mean water balance of the Tuul River at Ulaanbaatar station ......................................2 Table 2: Area of sub-catchments of the Tuul River.................................................................................3 Table 3: Percentage of runoff components ..............................................................................................3 Table 4: Annual average discharges of runoff components in m3 s-1.......................................................3 Table 5: Long-term mean values of meteorological variables (1965 - 2000)..........................................4 Table 6: Monthly average lumped rainfall in the Tuul River basin.........................................................4 Table 7: Data, instrument requirement and sources...............................................................................11 Table 8: Spatial and temporal information of water quality sampling points .......................................14 Table 9: Data collection and sources .....................................................................................................16 Table 10: Hourly average solar radiation in W h/m2 .............................................................................18 Table 11: Permissible level of surface water variables..........................................................................19 Table 12: Assessment of surface water quality......................................................................................21 Table 13: Definition for classification of surface water quality in Mongolia .......................................22 Table 14: Statistical summary of water quality index ...........................................................................23 Table 15: The quantity of samples with critical values .........................................................................23 Table 16: Annual mean water quality index ..........................................................................................31 Table 17: Initial condition in flow model ..............................................................................................41 Table 18: Statistical evaluation of sensitivity analysis ..........................................................................44 Table 19: Initial condition in quality model ..........................................................................................56 Table 20: Quality boundary condition ...................................................................................................56 Table 21: Statistical evaluation of very sensitive parameters................................................................60 Table 22: Statistical evaluation of less sensitive parameters.................................................................61 Table 23: Statistical evaluation of not sensitive parameters..................................................................61 Table 24: Scenario for SWQ improvement, Option 1 ...........................................................................63 Table 25: Scenario for SWQ improvement, Option 2 ...........................................................................64 Table 26: Scenario for SWQ improvement, Option 3 ...........................................................................64

viii

List of figures

Figure 1: Elevation map of the Ulaanbaatar city..................................................................................... 1 Figure 2: Climadiagram at the Ulaanbaatar station................................................................................. 5 Figure 3: Location of research field ........................................................................................................ 7 Figure 4: Methodology flowchart ......................................................................................................... 12 Figure 5: Photo at Terelj-Terelj sampling point.................................................................................... 13 Figure 6: Locations of sampling points................................................................................................. 14 Figure 7: Monthly mean water temperature.......................................................................................... 17 Figure 8: Monthly average wind speed at Ulaanbaatar station ............................................................. 17 Figure 9: Data frame of the Tuul River dataset..................................................................................... 23 Figure 10: Highest value of WQI.......................................................................................................... 23 Figure 11: Histogram and box plot of water quality index ................................................................... 24 Figure 12: Box plot for spatial variability of WQI................................................................................ 24 Figure 13: Correlation between distance & WQI.................................................................................. 25 Figure 14: Correlation between distance and WQI in the upstream section......................................... 26 Figure 15: Water quality fluctuation along the upstream in 1996-2006 ............................................... 26 Figure 16: Correlation between distance and WQI in the downstream section.................................... 27 Figure 17: Correlation between time & WQI (1996-2006)................................................................... 27 Figure 18: Water quality fluctuation at Tuul-Uubulan SP during study period.................................... 28 Figure 19: Water quality fluctuation at Tuul-Songino (upper) SP during study period ....................... 28 Figure 20: Water quality fluctuation at Tuul-Songino (down) SP during study time........................... 28 Figure 21: Water quality along the Tuul River in 1996 ........................................................................ 29 Figure 22: Water quality along the Tuul River in 2002 ........................................................................ 29 Figure 23: Water quality along the Tuul River in 2006 ........................................................................ 30 Figure 24: Seasonal water quality in upstream reach............................................................................ 30 Figure 25: Seasonal water quality in downstream reach....................................................................... 30 Figure 26: Correlation between river monthly mean discharge and monthly WQI.............................. 32 Figure 27: Flowchart of a flow model................................................................................................... 38 Figure 28: A flow model network ......................................................................................................... 39 Figure 29: Simplified cross-section of the Tuul River under Zaisan Bridge ........................................ 39 Figure 30: Simplified cross-section of the Tuul River under Altanbulag Bridge ................................. 40 Figure 31: Discharge of the Tuul River under Zaisan Bridge in 2005 and 2006.................................. 41 Figure 32: Artificial discharge at end node........................................................................................... 42 Figure 33: Correlation between calculated and observed discharge in calibration year....................... 43 Figure 34: Observed and modelled Q in calibration year ..................................................................... 43 Figure 35: Sensitivity analysis .............................................................................................................. 44 Figure 36: Correlation between calculated and observed Q data in validation year............................. 45 Figure 37: Observed and modelled Q in validation year....................................................................... 45 Figure 38: Schematization of DO balance model ................................................................................. 50 Figure 39: Flowchart of a quality model ............................................................................................... 54 Figure 40: Quality model network schematization ............................................................................... 54 Figure 41: Correlation between calculated and observed DO in calibration year ................................ 58

ix

Figure 42: Observed and modelled DO in calibration year ...................................................................59 Figure 43: Sensitivity analysis for beta..................................................................................................59 Figure 44: Sensitivity analysis for SOD ................................................................................................60 Figure 45: Sensitivity analysis for Vs....................................................................................................61 Figure 46: Correlation between calculated and observed Q data in validation year .............................62 Figure 47: Observed and modelled Q in validation year .......................................................................62 Figure 48: Options for SWQ improvement............................................................................................63

x

List of equations

Equation 1: Surface water quality index ............................................................................................... 21 Equation 2: Mass equation .................................................................................................................... 33 Equation 3: Momentum equation.......................................................................................................... 33 Equation 4: Discharge equation ............................................................................................................ 34 Equation 5: Correction factor for non-uniformity of velocity distribution........................................... 34 Equation 6: Hydraulic radius................................................................................................................. 34 Equation 7: Slope of riverbed................................................................................................................ 35 Equation 8: River velocity..................................................................................................................... 35 Equation 9: Froude number................................................................................................................... 35 Equation 10: Hydraulic depth ............................................................................................................... 35 Equation 11: Nash-Sutcliffe coefficient................................................................................................ 36 Equation 12: Mean error ....................................................................................................................... 37 Equation 13: Mean absolute error ......................................................................................................... 37 Equation 14: Root means squared error ................................................................................................ 37 Equation 15: Slope angle equation........................................................................................................ 40 Equation 16: Momentum balance equation........................................................................................... 48 Equation 17: Mass balance equation..................................................................................................... 49 Equation 18: Equation for the constituent transport by advection and dispersion ............................... 49 Equation 19: Background dispersion coefficient .................................................................................. 49 Equation 20: Shear stress equation ....................................................................................................... 49 Equation 21: Peclet number .................................................................................................................. 50 Equation 22: Dissolved oxygen balance ............................................................................................... 50 Equation 23: Temperature dependent oxygen mass-transfer velocity .................................................. 51 Equation 24: Re-aeration rate................................................................................................................ 51 Equation 25: Oxygen saturation............................................................................................................ 51 Equation 26: Primary production .......................................................................................................... 52 Equation 27: Oxygen consumption for sediment .................................................................................. 52 Equation 28: Ultimate BOD.................................................................................................................. 52 Equation 29: Biochemical oxygen demand........................................................................................... 53 Equation 30: Oxygen consumption for nitrification ............................................................................. 53 Equation 31: Dispersion coefficient...................................................................................................... 57

xi

List of appendices

Appendix 1: Photos in the field .............................................................................................................73 Appendix 2: Note for water field survey ...............................................................................................75 Appendix 3: Field measurements...........................................................................................................76 Appendix 4: Hydro-chemical dataset.....................................................................................................78 Appendix 5: Hydraulic dataset...............................................................................................................80 Appendix 6: CWTP chemical dataset ....................................................................................................82 Appendix 7: A flow model setup ...........................................................................................................83 Appendix 8: Manning value...................................................................................................................84 Appendix 9: Quality model syntax ........................................................................................................85 Appendix 10: A quality model setup .....................................................................................................86 Appendix 11: Required input data for the quality model ......................................................................87

xii

Abbreviations

ALAGC Administration of Land Affairs, Geodesy and Cartography

BOD5 Biochemical Oxygen Demand (5 day)

COD-Mn Chemical Oxygen Demand (Manganese III method)

CSM Centre of Standardization and Measurement

CWTP Central Wastewater Treatment Plant

DMS Duflow Modelling System

DO Dissolved Oxygen

ITC International Institute for Geo-Information Science and Earth Observation

IWEC International Weather for Energy Calculations

JICA Japanese International Cooperation Agency

MAE Mean absolute error

MASM Mongolian Administration of Standardization and Measurement

ME Mean error

MNE Ministry of Nature and Environment

NAMHEM National Agency of Meteorology, Hydrology and Environment Monitoring

NGIC for NRM National Geo-Information Centre for Natural Resource Management

NGO Non-governmental Organization

NSA National Standard Agency

NUM National University of Mongolia

RMSE Root mean squared error

SCLM State Central Library of Mongolia

SP Sampling point

SWQ Surface Water Quality

SWQI Surface Water Quality Indices

UB Ulaanbaatar city

WQA Water Quality Assessment

WQI Water Quality Index

WSSA Water supply and sewerage authority

WTS Wastewater Treatment Station

SURFACE WATER QUALITY ASSESSMENT AND MODELLING

INTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION 1

Chapter 1: Introduction

1.1: Background information

The background information of study area consists of four different parts.

1.1.1: Spatial information

Mongolia is located in the Central Asia between 41035’ - 52006’ North and 87047’ - 119057’ East.

Mongolian overall territory is 1.57 million square kilometres, total width is 1260 km from north to

south and total length is 2370 km from west to east.

Mongolia is situated in arid and semi-arid natural zone. The climate is harsh, with less

precipitation (approximately 200 mm year-1) and greatly fluctuating temperatures varying between

-350C in January and 320C in July.

The main type of land usage is, because of less rainfall, rangeland for nomadic livestock and

husbandry, fodder crop productions are of minor importance, the limiting factor is the lack of

water for irrigation. The total surface water resources in Mongolia is estimated 599 km3 year-1 and

is composed of water stored in lakes 500 km3 year-1, glaciers 62,9 km3 year-1 and rivers

34.6 km3 year-1 [Baasandorj and Davaa, 2006].

Ulaanbaatar, capital city of Mongolia, is placed in between E 106043’ - E 107002’ of longitude and

between N 47053’ - N 47057’ of latitude. Elevation ranges between 1214 m to 2079 m above mean

sea level and stretches from SW to NE.

Figure 1: Elevation map of the Ulaanbaatar city

Source: SRTM [SRTM, 2005]

A CASE STUDY IN THE TUUL RIVER, ULAANBAATAR CITY, MONGOLIA

DEPARTMENT OF WATER RESOURCE AND ENVIRONMENTAL MANAGEMENT 2

In territory of the Ulaanbaatar city, has about 49 streams and rivers (most of them are dried up).

Three rivers of them, which are named Selbe, Uliastai and Tuul, are running through the central

part of the Ulaanbaatar city. Two rivers are tributaries of the Tuul River, which is the biggest one.

The origins of the Tuul River are the Namiya and the Nergui streams at the southwestern slope of

the Khentii mountain range that is located in approximately 2000 meters above the mean sea level.

Terrain of the Tuul River basin ranges from 1200-2700 meters above mean sea level [Dashdeleg

and Bat, 1971].

The total length of the Tuul River is around 720 km and flows through the Ulaanbaatar city. Its

catchment area covers roughly 49,840 km2 (3.19% of the entire territory of Mongolia) and consists

of 11 soums of five aimags1 territories, which are namely Arhangai, Uvurhangai, Selenge, Bulgan

and Tuv. The Tuul River, located in the heart of the Ulaanbaatar city, is an environmentally,

economically and socially significant natural resource [Roza-Butler, 2004]. Almost hundreds of

tourist camps are to be found around the runoff formation zone and the gold mining activities are

taking place along the downstream of the Tuul River [Orgil, 2007].

1.1.2: Hydraulic information

Totally, eight hydrological stations were sited along the river basin. Four of them are stopped and

two of them are half-operational due to old technique and economic constraint. Rest two stations

are functioning, nowadays. The oldest hydrological station is “Tuul-Ulaanbaatar” which is

operating since 1942.

Table 1: Annual mean water balance of the Tuul River at Ulaanbaatar station

Precipitation 250 mm

Discharge 149 mm

Evapotranspiration 101 mm

Source: NAMHEM

The Tuul River basin can be divided into following eight sub-basins:

− Upper Tuul basin

− Terelj basin

− Khol basin

− Uliastai basin

− Selbe basin

− Turgen basin

− Middle-lower Tuul

− Kharuuh basin

Biggest tributary of the Tuul River in upper basin is the Terelj River that is draining from

catchment area of 1380.4 km2. After confluence of rivers Turgen and Tuul, there is nearly no

tributary in middle-lower Tuul sub-catchment. Size of sub-catchments area have been shown in

Table 2 [NAMHEM, 1999].

1 Soum and aimag are names of the administrative units in Mongolia



Table 2: Area of sub-catchments of the Tuul River

No Name of the sub-catchments Size of the catchment area (km2)

1 Upper Tuul 2698.5

2 Terelj 1380.4

3 Khol 249.5

4 Uliastai 308.9

5 Selbe 303.4

6 Turgen 531.6

7 Kharuukh 16576.2

Runoff components of the Terelj and the Tuul rivers were separated by hydrological method.

Hydrological and meteorological observations record that spring flood starts in middle of April and

low flow period occurs in June. Annual runoff of the Tuul River consists of the following three

contributors: (i) 69% from the rainfall, (ii) 26% from groundwater and (iii) 5% from snow. The

spatial distribution of groundwater contribution decreases along the Tuul River. This is because

80% of the Ulaanbaatar city’s water supply is provided by groundwater. Meanwhile, contribution

of precipitation increases in downstream with increment of catchment area [Baasandorj and

Davaa, 2006].

Table 3: Percentage of runoff components

Stations Year Groundwater Snow water Rainfall

Tuul – Bosgo bridge 2000 44.7 5.3 50.0

Terelj - Terelj 2000 41.6 5.2 53.3

Tuul – Ulaanbaatar 2000 37.4 7.3 55.2

The average channel width of the Tuul River is 35 to 75 meters during non-flooding time, depth is

0.8-3.5 m and the velocity is 0.5-1.5 m s-1. The annual mean flow of the Tuul River is

approximately 26.6 m3 s-1. The observed maximum discharge reached to 1580 m3 s-1 and 564 m3 s-1

in Ulaanbaatar and Terelj stations, respectively. During the low flow period of warm season, it has

dropped until 1.86 m3 s-1 at Ulaanbaatar and 0.44 m3 s-1 at Terelj stations [NAMHEM, 1999].

Table 4: Annual average discharges of runoff components in m3 s-1

Stations Year Groundwater Snow water Rainfall Annual mean flow

Tuul – Bosgo bridge 2000 6.06 0.73 6.78 13.6

Terelj - Terelj 2000 2.70 0.34 6.08 9.12

Tuul – Ulaanbaatar 2000 4.90 0.96 7.23 13.1

Water demand of the city has increased by 20% from 1998 to 2005. Population growth,

urbanization and intensity of industries have created water exploitation, deterioration of natural

water regime and ecological degradation of the Tuul River basin. Moreover, the Tuul River drains

into the Orhon River, one of the main tributaries of the Selenge River. The Selenge River is the

main tributary of the Lake Baikal in Russia, the world largest freshwater lake by volume. The Tuul

River basin is economically most important and one of the mainly polluted rivers in Mongolia.

However, no management plan currently exists for the water resources of the Tuul River basin

[Orgil, 2007].



1.1.3: Climatic information

In the Tuul River basin, has continental climatic feature that is characterized by wide variations of

annual, monthly and daily temperatures, low range of air humidity, non-uniform distribution of

precipitation, cold, long-lasting winter and warm summer. The rainy period continues from June to

August in the upper Tuul River basin, of which rainfall shares about 74% of the annual

precipitation.

Annual average air temperature is -1.20C in the study area. Annual minimum temperature is

recorded -39.60C in January, maximum temperature reaches 34.50C. Fluctuation of air

temperatures reaches approximately 74.10C.

Table 5: Long-term mean values of meteorological variables (1965 - 2000)

Variables Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Annual

Precipitation, mm 2.2 1.6 3.7 7.3 14.7 54.6 57.9 75.9 23.4 9.7 4.2 3.2 258.5

Air temperature, 0C -21.8 -18.1 -8.8 0.8 9.4 14.9 16.9 14.8 8.3 0.1 -11.4 -19.5 -1.2

Wind speed, m s-1 1.6 2.1 2.9 3.9 4.4 3.5 3.1 2.6 2.9 2.6 2.0 1.6 2.8

Annual mean precipitation is 258.5 mm at Ulaanbaatar station and almost 90 percent occurs in

warm session of year, particularly in April to September. The mean value of precipitation, which

occurs in warm period, is 233.8 mm that is falling dominantly in form of thunderstorm. Daily

maximum precipitation occurred in 1967 that recorded 74.9 mm [Baasandorj and Davaa, 2006].

Table 6: Monthly average lumped rainfall in the Tuul River basin

Stations Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Cumulated

Buyant-Uhaa 1.48 1.40 2.3 6.29 13.72 43.29 68.00 73.73 32.66 9.15 6.82 2.85 258.5

Tahilt 1.79 1.43 2.64 7.59 13.40 48.77 69.90 76.20 32.78 8.69 4.32 3.07 270.5

MUIS 1.93 1.64 4.54 9.28 14.22 49.00 72.04 75.54 31.30 9.62 4.87 2.92 276.8

In the study area, wind direction dominants from north and northwestern. However, surrounding

mountains make change in locally. Annual mean wind speed is recorded 2.8 m s-1. The 53.6% of

annual wind speed falls in range of 0-2 m s-1, 26.8% falls 2-5 m s-1, 12.5% belongs to 6-10 m s-1

and only 1.3% concerns to the 11-15 m s-1. Annual air humidity is 62% and its range in December

and January is 70-74%, lowest value is 48% in April. Average vapour pressure is 4.8 and

saturation deficit is 3.8 hPa. Snow in the study area covers 68 days in average. It begins from

October and stabilizes in middle of November, melts in May [Baasandorj and Davaa, 2006].

Climadiagram of study area at Ulaanbaatar meteorological station has been shown by the

following graph.



Figure 2: Climadiagram at the Ulaanbaatar station

-30

-20

-10

0

10

20

30

40

50

Jan-

04F

eb-0

4M

ar-0

4A

pr-0

4M

ay-0

4Ju

n-04

Jul-0

4A

ug-0

4S

ep-0

4O

ct-0

4N

ov-0

4D

ec-0

4Ja

n-05

Feb

-05

Mar

-05

Apr

-05

May

-05

Jun-

05Ju

l-05

Aug

-05

Sep

-05

Oct

-05

Nov

-05

Dec

-05

Jan-

06F

eb-0

6M

ar-0

6A

pr-0

6M

ay-0

6Ju

n-06

Jul-0

6A

ug-0

6S

ep-0

6O

ct-0

6N

ov-0

6D

ec-0

6

0C

-60

-40

-20

0

20

40

60

80

100

mm

Air temperature Precipitation

Source: Data from NAMHEM

1.1.4: Hydro-chemical information

The Tuul River quality is naturally clean and rich of calcium, bicarbonate. Total dissolved solid of

the river water ranges from 100-210 mg l-1 along its reaches. The Tuul River contains 78.9% Ca+2,

15.8% Na+, 28.1-634.7 mg l-1 mineral and it belongs to the hydro-carbonate class, calcium group,

pH = 6.1-7.5. The main cation is calcium and dominant anion is hydro-carbonate in the Tuul River.

Moreover, cation proportion is Ca+2 > Mg+2 > Na+ + K+ and anion ratio is HCO3- > SO4

-2 > CI-.

Naturally, anion and cation proportions and chemical content of water matches to the river with

pure water [NAMHEM, 1999].

Along the Tuul River, water quality has been monitored since 1980th. Chemical and biological

variables such as pH, nitrite, nitrate, ammonium, phosphate, DO, COD-Mn, and BOD5 are

analysing in the CLEM. The Tuul River is nearly not polluted until the Ulaanbaatar city and

pollution starts during running through UB. Most of tributaries, especially rural areas, are not

affected by human activities and have good self-purification rate [Baasandorj and Davaa, 2006].

Unfortunately, in the last few years, rapid urbanization and increment of industries have negatively

influenced water quality and chemical composition of river in surrounding area of the Ulaanbaatar

city [Javzan, et al., 2004]. Therefore, chemical content of river suddenly changes from west part of

the Ulaanbaatar city. The reason of chemical changes is the non-completely treated wastewater

from the Central Wastewater Treatment Plant (CWTP), which is located in west side of the

Ulaanbaatar city, pours into the Tuul River [Altansukh, 2000].

1.2: Recent problem statement

Water quality and pollution of the Tuul River has been monitored in 10 stationary points along the

river since 1980s. Fast urbanization combinations with increasing number of tourist camps,

agricultural and mining activities have had a significant negative impact on the Tuul River’s quality

and its associated ecosystems. Consequently, water becomes seriously polluted, loses its clarity,

transparency and self-purification rate decreases in year by year.

According to results of the hydrological survey was conducted in 2003, hydrological regime and its

runoff formation zones of the Tuul River are gradually being changed and polluted by the settlements,



intensive overgrazing, timbering, wild fires and improper wastewater treatment in the river banks

[Baasandorj and Davaa, 2006].

Water contamination of the Tuul River is continuously increasing from upper to its lower reach.

Naturally, upstream of the Tuul river is running through mountainously area and there has high

velocity and turbulent. Hence, upper part of river has more oxidization potential, re-aeration and self-

purification. When it comes to the Ulaanbaatar city, natural condition changes from mountainous

region to valley. In valley, velocity and turbulence of river decreases, then capability of oxidization,

re-aeration and self-purification also reduces as well. This is the natural factor of possibility to store

contaminant elements in the river water a longer time and distance [Altansukh, 2000].

In last decade, contaminant of the Tuul River is constantly increasing related to increment of

industries, agricultures and old, insufficient sewage and treatment system. Along the Tuul River has

five contamination point sources. The self-purification coefficient of the Tuul River is 6.57 until it

reaches first contamination point source and it reduces until 0.98 after the CWTP wastewater is

pouring into river. The main and biggest artificial point source of pollution in the Tuul River basin is

improper treated wastewater from the CWTP (190,000 m3 day-1), then Nalaih (1,400 m3 day-1), Niseh

(400 m3 day-1), Bio-industry (490 m3 day-1) and Bio-Songino wastewater treatment station (600 m3

day-1). The CWTP is responsible for the collection and treatment of industrial and domestic sewage

water in Ulaanbaatar. The treatment efficiency of the CWTP as well as other wastewater treatment

stations in the region is often inadequate due to financial constraint. The efficiency of the CWTP is

approximately 60-70% due to poor maintenance, lack of spare parts, outdated equipment and frequent

power shortages. This also causes of 10-20% wastewater on a daily basis to be released directly into

the Tuul River without any treatment [Orchlon, 1995]. Air, soil pollution and accumulated wastes in

catchment area, which are transferred by surface runoff and flood channel, are also have significant

impact on river water quality. Population of the Ulaanbaatar city is produced around 800-1000 tons of

dry wastes per day. Efficiency of the CWTP was 71% in 2002. This percentage has dropped to 66%

in 2003 and treatment level was lower than 50%, even sometimes was not operated in May 2003 and

April 2004. The major causes of water pollutant are mining industries in lower basin of the Tuul

River. Approximately, 179 licensed mining companies are operating in 145 km2 area of the basin

[MNE, 2006].

Stationary hydro-biological monitoring of invertebrate species along the Tuul River has started since

1997. Aquatic and native communities have disappeared due to pollution and gold mining activities

[MNE, 2006].

1.3: Reason of topic selection

Based on the following motivations, this topic selected for an MSc research.

− Data availability

− Attention from public and interest groups

− Current situation

− Research of the NGIC for NRM project, MNE

− Possibility of future study

The hydro-chemical and hydraulic data are availably, which measures in the Central Laboratory of

Environmental Monitoring (CLEM) in each month and National Agency of Meteorology, Hydrology

and Environment Monitoring in daily, respectively. However, there is no thematic map of surface



water contaminant. Therefore, this kind of research is still open in Mongolia and it has opportunity in

future study using different contaminant variables, in different location.

Moreover, the Tuul River is most contaminated river in Mongolia and it is running through capital

city, utilizes for manufacture and drinking purposes. That is why; the public, political groups

including NGO and scientific groups pay attention to this river.

This research closely related to the water management sub-component of the NGIC for NRM project,

MNE that is ongoing in Mongolia. In addition, Dutch government has funded this project.

Therefore, the Tuul River provides an appropriate case for this study.

Moreover, as the availability of more powerful digital computers, modelling techniques and software

has rapidly increased, the use of both physical and analogue models in hydrology has been largely

replaced by that of computer implemented mathematical models, which are usually cheaper and much

more flexible [Dingman, 2002].

1.4: Research object

Study objects are the Tuul River and its three tributaries, namely Terelj, Uliastai and Selbe Rivers,

which are running through the capital city of Mongolia.

Figure 3: Location of research field

1.5: Research targets

− River quality assessment and pollution mapping in above mentioned rivers (total 14 sampling

points) using nitrite NO2-, nitrate NO3

-, ammonium NH4+, BOD5, COD-Mn, DO (total 6

chemical variables)

− River flow and quality modelling using hydraulic variables (discharge, level), physical

parameter (cross-section of river channel, location, distance), climate data (wind, solar

radiation) and chemical variables (NH4+, BOD5, DO)

1.6: Previuos studies

Previous studies about the Tuul River can be divided into the following two different directions:

1. Quality studies using chemical method

2. Quality studies using biological method

#

##

##

#

##

#

##

#

#

#

Òóóë - Õà äàíõ ÿñàà

Òóóë - Àëòàí áó ë àã

Òóóë - Øóâ óó í ôàáðè ê

Íà ëàéõ

Ãà÷ óóðò

Ñî íãèí î

Óëààíáààòàð

Òóóë - Çà éñàí

Ñýë áý - Óëààíáààò àð

Òóóë - Ñîí ñã îë îí

Òóóë - Ñî íãè í û ä îî ä

Òóóë - Ñîíãèí û äýýä

Òýð ýëæ - Òýðýëæ

Òóóë - Óóáóë àí

Òóóë - Íà ë àé õ

Òóóë - Áàÿíç óð õ

Óëè àñòàé - Óë àà íáà àòàð

Äó íä ã îë - Óë àà íáàà òà ð

Í àð èé íè é ãîë

Áóð

ä ãîë

Íà ðèé íèé ãîë

Íà ðè éí èé ã îë Ìàíäàë

ûí ãîë

Õóéí

ãî ë Øàâà ðò ãîë

Õóéí ã

îë

Øèæ èðèéí ãî ë

Òàõ èë

ò ûí àð ãîë

Ò îëãî éòû í ãî ë

Ñýë áý ã

îë

Õàíäãà éòûí ã

î ë

Ñ ýëáý ã

îë

Áèë ã

ý õèé í ã

îë

Ñ ýëõ ã îë

Áèëãý õè

é í ãî ë

Ñ ýëáý ãîë

Äóíä ãî ë

Áàðó óí ñà ëàà ã îë

Çóó

í ñ àëàà

ãîë

Ò îëãîéò

ûí ãîë

Äóíä

ã î ë

Òàõèë

òûí ã

îëÁàÿí ãîë

Áàð óó í ãî ë

Óëèà ñ

ò àé í ãîë

Ç óóí ã

îë

Õîéä áà

ÿí ãîë

Ó ðä áàÿí ã îë

Óëèà ñ

ò àé í ã

îë

Óëè àñ

òàéí

ãîë

Õà÷

óóð ãîë

Áó

ãàòû í ãî ë

Øàâàð

òûí ã

îë

Øèëæèð ãîë

Ãà÷óó

ðòûí

ãîë

Èõ õó

àí äûí ã

îë

Äýí ä è

éí ãîë

Òý ðý ëæè

é í ã îë

Òýðý

ëæèéí ãîë

Òý ðý ëæ èéí ãîë

Ò ýðý ëæè éí ãîë

À ð ãîð

õè

Òóóë ãîë

Òóóë ã

îë

Ò ó ó

ë ãîë

Ýëñòè éí ãîë

Ýëñò

ãîë

Ýëñò ãî ë

Îâ îð ãîðõ è

Íàëà

éõûí

ãîë

Ò

ó óë ã îë

Õîë è

éí ãî

ë

Òî ð õóð

õû í ãîë

Á óìá àò

û í ã îð õè

Õîëè é í ãîë

Ò óó ë ãî ë

Ò ó ðã ýí èé ã

îë

Îâîð ç à

é ñà íãèéí ã

îë

Òóðãý íèé

ãîë

Áóõýã ãîë

Òóó ë ãîë

Ò óóë ãîë

Òóóë ãîë

Òóóë ãîë

Territory of th eTehnogen are aWater pollutio nWater pollutio nWater pollutio nStudying river .Unstudying ri v

# Research poi n



1.6.1: Quality studies using chemical method

− The first sample for chemical analysis of the Tuul river was collected in July, 1945 by

Russian research team [Tuvaanjav, 1983].

− Around 1950, a Russian researcher N.T.Kuznetsov was done a study for chemical component

of the Tuul River and concluded total mineralization not exceeds 200 mg l-1, hydro-carbonate,

sodium, potassium ions are predominant and chlorine, sulphate ions are minor [Tuvaanjav,

1972].

− Tuvaanjav, G in 1978, a researcher was done chemical analysis in the Tuul river and its

tributaries and found out the following conclusions; the Tuul river is belongs to hydro-

carbonate class and contains 78.9% calcium, 15.8% sodium, rest of percent is magnesium,

total mineralization is between 28.1-634.7 mg l-1. Upper reach of river and its tributaries

contain high amount of sulphate than chlorine, but contrary in lower reach. Because of, urban

area has negatively affects to river. Mineralization, pH and ammonia concentration increase

along the Selbe and the Uliastai Rivers [Tuvaanjav, 1978].

− In a research report written by Munguntsetseg, A and et all, they were estimated that the river

is totally self-purified along 170 km of downstream after wastewater from CWTP pours into

river [Munguntsetseg, et al., 1982].

− An article from Tuvaangav, G, the 40 million cubic meter water used in overall demand of

capital city and it strongly affects to natural source of water near by city [Tuvaanjav, 1983].

− A researcher Davaa, G in 1996 was done study in 12 points along the Tuul River and its 4

tributaries and noted that total mineralization increases along the river, because of, soil type

changes, precipitation decreases and evaporation increases [MNE, 1997a].

1.6.2: Quality studies using biological method

− By M. Kolikvitts and D.Marson in 1908-1909, they made the first attempt to assessing water

quality using characteristics of water insects that sensitive in living condition. After this

attempt, many scientists were trying to assess water quality using biological method and are

developing new methods in recent years [Batima, 1998].

− In 1930, the bacteriological study in river and spring water was done by A.Kharit, Russian

scientist. A scientist was trying to find relation between water bacteriology and stomach

illness [Adiyabadam, 1996].

− A research of aqua insects has done by I.M.Levanidova in 1947, a study about horsefly fauna

implemented by B.P.Belisheva, A.Dashdorj in 1958 and other aqua insects was studied by

Mongolian and Russian scientific team in 1972 and 1979, respectively. They estimated fauna

of aqua insects in surrounding area of the Ulaanbaatar city [Soninhishig, 1998].

− K.Shenon, E.A.Chebotarev and D.Lenat, biologists, were developed their methods for water

quality in 1948, 1986 and 1990, respectively [Zagas, 1998].

− A new method for water quality using water worm developed by Michigan and others in 1955

[Adiyabadam, 1996].

− In between 1990-1997, Institute of Biology and Institute of Geo-ecology of Mongolian

Academy of Sciences was cooperated with Institute of Water, MNE were done hydro-

biological research in the Tuul River and estimated 170 aqua insects. Moreover, team

assessed water quality using those fauna [Munguntsetseg, 1987].



− N.Soninhishig in 1998, she defended an MSc degree in Mongolia by thesis title “Algae in the

Tuul River”. She was exploring algae fauna in river and was assessed the following water

quality index by algae fauna [Soninhishig, 1998].

� Totally, 47 algae species found under Terelj Bridge of the Tuul River and 20 of them

live in natural pure water, 5 of them live in contaminated water; rest is not indicator

of water quality. She calculated quality index and river water belonged to first class.

Her conclusion was that river reach is pure.

� Entirely, 29 species found at confluence of the Khol and the Tuul rivers and nine of

them live in natural pure water, 5 of them live in contaminated water; rest is not

indicator of water quality. She calculated quality index, which belongs to the third

class and concluded that section of river is slightly polluted.

� Thirty sex species found under Bayanzurh Bridge, 13 were indicator of pure water, 7

were indicator of polluted water and concluded that; quality index is two, which

means, comparatively clean water.

� At the Songino (upper) study point was polluted because of quality index was four.

� At the Songino (down) stationary sampling point was heavily polluted, as a result of

quality index five.

� At the Chicken farm, river water assessed heavily polluted and quality index belonged

to fifth class.

� At the Khadanhyasaa and Altanbulag, stationary sampling points were dropped in

quality index four.

− An MSc thesis from B.Zagas and Ya.Oyunchuluun, both of them were used riverbed insects

for assessment of quality and concluded same results with previous study in 1998 and 1999,

respectively [Zagas, 1998], [Oyunchuluun, 1999].

− In 2000, Uranbileg, L set zones of river water quality in surrounding area of Ulaanbaatar

using biological method by Lenat, D, American scientist. She estimated totally four zones of

river water quality [Uranbileg, 2000].

Recent time, water quality is studying by several organization and researchers. Nowadays, there have

four institutes are concerning water issues, continuously, namely Central Laboratory of

Environmental Monitoring, Water sector of Institute of Geo-ecology, Water sector of National

Agency of Meteorology, Hydrology and Environmental Monitoring, Water Authority of Mongolia

and universities. In addition, number of projects are implementing in water issues and some of them

funded by Dutch government.

1.7: Research objectives

Research objective generated from practical problem in Mongolia.

− To assess river water quality using SWQ indices which based on nitrite NO2-, nitrate NO3

-,

ammonium NH4+, BOD5, COD-Mn and DO in case study of the Tuul River near by

Ulaanbaatar city

− To develop a prototype water quality model of the Tuul River using the Duflow Modelling

System



The following research tasks are developed from research objectives.

− Fieldwork, collecting samples, estimate location of point and non-point contamination sources

and collect existing hydraulic, hydro-chemical datasets

− Analyze water samples in laboratory

− Integrate the hydro-chemical data of selected rivers in last decade near to the Ulaanbaatar city

using availably data

− Assess river water quality using combination of nitrite, nitrate, ammonium, BOD5, COD-Mn,

DO and utilizing surface water quality indices

− Determine spatial and temporal changes of water quality

− Visualize thematic maps of river water contamination

− Develop a Tuul river flow model using DMS

− Build up a surface water quality model based on a flow model

1.8: Research questions

The following questions arose during the research and tried to find out answers:

− Which methods are suitable for our research? (standard methods or state-of-the-art)

− What kind of change was occurred in study period? (negative or positive)

− What type of model can we develop? (lumped, distributed and steady-state, continuous)

− Is DMS suitable for simulating flow and quality models of selected river? (yes or no)

− Is the output of research valuable and significant? (yes or no)

− If yes, that significance for whom? (local, regional and personal, groups)

1.9: Research phases

Entire period of this research divided in three main phases.

1) Pre-field work

− To review literatures

− To select methods and required data

− To prepare fieldwork

− To understand how Duflow modelling system works

2) During field work

− To collect samples and analyze in laboratory

− To gather availably datasets of last decade (1996 - 2006)

− To identify point and non-point sources of water contamination

− To collect a topographic map of study area

3) Post-field work

− To assess and classify river pollution using hydro-chemical datasets and SWQI

− To digitize topographical map of study area

− To estimate spatial and temporal changes of water quality

− To visualize thematic maps of river contamination

− To develop a river flow and a water quality model

− To set up a scenario for SWQ improvement in quality model



Table 7: Data, instrument requirement and sources

N Data Source

1 Literature ITC, internet, SCLM, Academic library of NUM and others

2 Sample from field

3 Dataset CLEM, NAMNEM and others

4 Map ALAGC

5 Documents (standard, book) MNE, NGIC project, MASM

6 Field equipment ITC

7 Computer ITC

8 Software ITC

9 Laboratory CLEM

1.10: Methodology

In this research, the following methodologies are selected, including:

− Literature review

This stage was encompassed the search and review of publications related with water quality

assessment, pollution mapping and water quality modelling issues in the study area and it is

important for estimate the research objectives.

− Field survey

This phase was involved measurement of chemical variables for water quality assessment of

the Tuul River and general survey of study area. The pH, EC, DO variables measured in the

field and rest of variables analyzed in the CLEM. Water samples are collected by standard

method MNS ISO 5667-6:2001 “Water quality. Sampling, 6th part. Guideline for sampling

from river and stream”. Therefore, contamination sources identified during field survey.

− Laboratory analysis

The chemical variables NO2-, NO3

-, NH4+ and COD-Mn were analyzed by spectrometer

method in the CLEM using HACH instrument. BOD5 was measured by dilution method in

same laboratory.

− Data collection

This step consisted of gathering the available data such as topographic map, hydro-chemical

datasets from previous years, hydraulic variable, physical parameter and climatic data from

different organization in Mongolia.

− Data analysis

In this stage, all data, which are collected and measured during fieldwork, assessed and

classified using SWQI

− Map visualization

Thematic maps for spatial and temporal changes of SWQ visualized using ILWIS software.

− Model development

A flow and a quality models were developed using Duflow Modelling System.

More than one method and technique were used to attain one task.



Figure 4: Methodology flowchart

1.11: Research hypothesis

Obviously, the contamination of the Tuul River near by the Ulaanbaatar city has been continuously

increasing during last decade related to the increment of industrialization, agricultures and old,

insufficient sewage, treatment system.

It is quite possible to estimate temporal and spatial changes of water quality. Furthermore, according

to the User guide of Duflow software, it can develop a model of surface water quality using physical

parameter, hydrological variables, chemical variables and climatic data. Based upon that model, we

can set up a scenario and predict the effect of contamination source in water quality in the future.

1.12: Outline of the thesis

Background information of study field, the objectives of the research, recent problem statement,

previous studies in that field and methodology are included in the main subjects of chapter 1. Chapter

2 provides information about the selection of sampling site, data collection and field survey. The

following chapter 3, which is named water quality assessment, is mainly focused on spatial and

temporal assessment of surface water quality using geo-statistical tools and surface water quality

index. The thematic maps of surface water quality presented in chapter 3. Chapter 4 concentrated for

the Tuul River a flow model using Duflow modelling system. Process and results of a quality model

for DO in case of the Tuul River is mentioned in chapter 5. Chapter 6 provides the overall

conclusions, recommendations from this research and future studies.



Chapter 2: Field survey and data collection

2.1: Field survey

The research field campaign was executed during 2 months, starting from 18 August until 24 October,

in the Ulaanbaatar city, Mongolia. Approximately, 800 km was travelled (including daily sample

transportation) during fieldwork and totally, 14 sample points were investigated for hydro-chemical

variables, 55 sites selected for measurement of general variables. HACH 40D field instruments and

ETREX hand-held GPS was used in the field campaign. Totally, 52 photos were taken in the field at

each site.

2.1.1: Selection of sampling site

According to classification of rivers based on discharge, drainage area and river width which

written in book by Chapman, D [WHO, 1996], the Tuul River is a small river. Because of, its

channel average width 35-75 meter during non-flooding time, the depth is 0.8-3.5 meter and the

velocity is 0.5-1.5 m s-1. The annual average discharge is approximately 26.6 m3 s-1. Drainage area

is in the region of 49,840 km2 and total length is 720 km, it flows generally in western direction.

Consistent with Mongolian river classification, developed by Davaa, G, which is based on long-

term annual mean flow, the Tuul River is a moderately big river.

CLEM chose the following 10 sampling points along the Tuul River and 4 sites in tributaries of the

Tuul River, one in Terelj, one in Uliastai, two in Selbe, in 1980th. Water quality and pollution in

surrounding area of the UB city has been monitored in those 14 points since 1980s.

During the field campaign, 14 samples collected from those sites for ammonium, nitrate, nitrite,

BOD5, COD-Mn examination and analyzed in the CLEM. In addition, some photos were captured

in the field. See some representative photos in figure 5 and Appendix 1.

Figure 5: Photo at Terelj-Terelj sampling point



The following a table and a map show the geographical locations of sampling points.

Table 8: Spatial and temporal information of water quality sampling points

N Name of sites Latitude Longitude Altitude

in m

Temporal

sampling Selection

1 Terelj - Terelj 47°59'30.67"N 107°27'35.55"E 1522 monthly Tributary of main river

2 Tuul - Uubulan 47°48'26.40"N 107°22'50.30"E 1383 monthly Base load

3 Tuul - Nalaih 47°49'14.00"N 107°15'56.40"E 1364 monthly Discharge from local WTS

4 Tuul - Bayanzurh 47°53'28.10"N 107°03'04.70"E 1309 monthly Inflow to city

5 Tuul - Zaisan 47°53'19.40"N 106°55'05.70"E 1293 monthly Centre of city

6 Tuul - Sonsgolon 47°52'28.70"N 106°46'50.10"E 1272 monthly Outflow from city

7 Tuul - Songino (upper) 47°51'17.80"N 106°41'23.20"E 1256 monthly Upper reach of CWTP

8 Tuul - Songino (down) 47°50'51.70"N 106°40'29.70"E 1254 monthly Lower reach of CWTP

9 Tuul – Chicken farm 47°46'21.00"N 106°35'59.20"E 1233 monthly Discharge from bio-industry

10 Tuul - Khadanhyasaa 47°45'08.90"N 106°30'02.60"E 1217 monthly Indicator of self-purification

and inflow to town

11 Tuul - Altanbulag 47°41'53.40"N 106°17'40.60"E 1182 monthly Indicator of self-purification

12 Uliastai - UB 47°54'07.80"N 107°01'51.77"E 1310 monthly Tributary of main river

13 Selbe - UB 47°54'30.77"N 106°55'55.77"E 1290 monthly Tributary of main river

14 Dund - UB 47°54'11.96"N 106°51'23.25"E 1276 monthly Tributary of main river

Hint: Geographical coordinate and altitude are measured by hand-held GPS, Etrex.

Figure 6: Locations of sampling points

There are five contamination point sources, which are represented by red triangles and 14 yellow

diamonds point out the sites where water samples were taken for above-mentioned five hydro-

chemical variables. In addition, pink dots are indicating the location of in-situ measurement places

where were measured general variables t0, pH, EC and DO using the HACH 40D field instrument for

water survey. Note: some of them might be coinciding.



2.1.2: Sampling method

Surface water can be sampled using isokinetic and non-isokinetic samplers. Isokinetic samplers are

made such that stream water approaching and entering the sampler intake does not change in

velocity, and are consequently more expensive and harder to operate. In this research, we are used

different types of non-isokinetic samplers, namely open-mouth samplers. The hand-held plastic

bottle sampler is the simplest type of open-mouth sampler: a sample bottle is used as sampler. To

sample water from a river, hold the bottle mid of water depth, with the mouth facing slightly

downward. Turn the bottle upright to fill it and replace the bottle cap. [Dost, 2006].

2.1.3: Chemical analysis in field

A number of variables, pH, EC and DO must be measured in-situ very soon after the sample has

been collected because the value of these variables either change rapidly after sample collection

[Dost, 2006].

In the field, general variables t0, pH, EC and DO measured at least two times using HACH 40D a

field instrument. First, took water sample into a plastic container and measured by an instrument.

Secondly, probes immersed into flowing water and measured again. Difference of measurement

results was small and calculated average value.

Moreover, a field survey note prepared before the field campaign and totally, 55 sheets were filled

in-situ. Field survey sheet contains three clusters of information:

− Information of sampling location

− Sample information

− Measurement results

And, it includes the following information such as name of water body, basin, sub-basin,

geographical coordinate, elevation, land use around sampling point, possible contamination source,

name of sample collector, collected date, sample ID, sampling method, preservation method,

purpose of sampling, weather condition, any changes in weather condition, variable measured,

actual result, measured unit, analytical method and etc. For more information, see Appendix 2.

Beside of this, general variables were measured at 55 sites along the river. Aim of field

investigation was to estimate fluctuation of DO, pH, EC along river, identification of pollution

sources, and surveillance of entire study field. For the field measurements, see Appendix 3.

Brief steps of in-situ measurement:

− Prepare instrument

− Take a sample from river using a plastic container

− Measure variables

− Write and save results

− Immerse probes in river

− Measure variables, again

− Compare with previous result

− Calculate average number

− Fill field note

In situ measurement steps that are more detailed look at HACH 40D user manual or field guide for

water quality sampling and testing [Dost, 2006].

Chemical variables ammonium, nitrite, nitrate, COD-Mn and BOD5 in this study was measured in

the standard laboratory, CLEM in Ulaanbaatar, within a day after the sample had been collected



from above-mentioned sampling points. These hydro-chemical variables are unstable and

maximum permissible storage time is 48 hours.

In the field, 0.5-litre samples collected for laboratory analysis from river using plastic container,

were kept in a storage box and sent it to laboratory, analyzed by DR 2010 spectrophotometer

within a day. No preservative was added into samples. For more information about analysis

methods, see the procedure manual of the DR 2010 instrument.

2.2: Data collection

Beside of above-mentioned survey, some existing hydro-chemical, hydraulic and climatic datasets are

collected during field campaign. The following table has shown data collection from different

organization.

Table 9: Data collection and sources

N Data Implemented organization Used for

1 Hydro-chemical data CLEM Assessment, mapping, modelling

2 Hydraulic data WS, NAMHEM Modelling

3 Climatic data NAMHEM Modelling

4 Map and standards ALAGC and CSM Assessment, mapping, modelling

5 Pollution source CWTP and WSSA Assessment, modelling

2.2.1: Chemical data

The following chemical variables are included in hydro-chemical datasets:

− DO dissolved oxygen

− NH4+ ammonium

− NO2- nitrite

− NO3- nitrate

− COD-Mn chemical oxygen demand

− BOD5 biochemical oxygen demand

− T0water water temperature

Chemical dataset covers totally 11 years, 1996-2006. Water samples were taken from above-

mentioned 14 sampling points and examined in the CLEM. For an example of dataset, see

Appendix 4. All other hydro-chemical datasets prepared same as in Appendix 4.

Chemical dataset was used to assess quality of the Tuul River and its tributaries in surrounding

area of the UB city. Several quality distribution maps visualized by ILWIS based on assessment.

Moreover, DO, NH4+, BOD5 and water temperature data are input of a quality model.



Figure 7: Monthly mean water temperature

0.00

2.00

4.00

6.00

8.0010.00

12.00

14.00

16.00

18.00

Apr-05 May-05 Jun-05 Jul-05 Aug-05 Sep-05 Oct-05 Nov-05

Month

Wat

er te

mp

erat

ure

in C

2.2.2: Hydraulic data

Determining the hydrological regime of a water body is an important aspect of water quality

assessment. Discharge measurements, for example, are necessary for flow or mass balance

calculations and as inputs for water quality models [WHO, 1996].

Water Sector of NAMHEM has measured all of hydraulic variables such as velocity, discharge,

river width and depth, which were utilized this research. Hydraulic data, which was measured in

2005 and 2006, was used in river flow modelling. For hydraulic dataset, see Appendix 5.

Furthermore, daily discharge was gauged at Tuul-Ulaanbaatar and Tuul-Altanbulag stations,

respectively.

2.2.3: Climatic data

Climatic dataset includes hourly average solar radiation and monthly mean wind speed of the study

area. This dataset measured by NAMHEM at the Ulaanbaatar station, world weather station code

is 442920, and obtained from online open source IWEC. Downloaded wind speed and solar

radiation datasets are shown below in Figure 8 and Table 10;

Figure 8: Monthly average wind speed at Ulaanbaatar station

0

0.5

1

1.5

2

2.5

3

3.5

4

Month

Win

d sp

eed

in m

/s

Wind speed 1.4 1.8 3.2 2.9 3.7 3.6 3.2 3 2.8 3 2.6 1.6

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec



Table 10: Hourly average solar radiation in W h/m2

Time Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

0:01- 1:00 0 0 0 0 0 0 0 0 0 0 0 0

1:01- 2:00 0 0 0 0 0 0 0 0 0 0 0 0

2:01- 3:00 0 0 0 0 0 0 0 0 0 0 0 0

3:01- 4:00 0 0 0 0 0 0 0 0 0 0 0 0

4:01- 5:00 0 0 0 0 0 0 0 0 0 0 0 0

5:01- 6:00 0 0 0 0 0 11 0 0 0 0 0 0

6:01- 7:00 0 0 0 69 135 155 103 33 0 0 0 0

7:01- 8:00 0 0 63 282 353 329 210 177 127 6 0 0

8:01- 9:00 0 84 287 542 472 422 318 312 304 172 61 0

9:01-10:00 192 325 475 704 479 461 385 382 420 384 243 160

10:01-11:00 367 543 523 627 468 460 411 398 456 502 399 381

11:01-12:00 517 594 585 600 477 466 389 382 468 555 488 488

12:01-13:00 541 564 596 552 461 464 344 346 441 558 520 492

13:01-14:00 534 513 567 441 432 452 290 308 408 522 513 449

14:01-15:00 509 514 560 502 405 439 290 348 430 488 465 434

15:01-16:00 423 481 531 576 366 412 283 379 415 395 334 316

16:01-17:00 203 362 458 442 330 375 270 379 354 225 95 38

17:01-18:00 0 140 354 463 307 339 276 322 258 39 0 0

18:01-19:00 0 0 55 310 210 264 219 203 54 0 0 0

19:01-20:00 0 0 0 6 48 128 86 19 0 0 0 0

20:01-21:00 0 0 0 0 0 0 0 0 0 0 0 0

21:01-22:00 0 0 0 0 0 0 0 0 0 0 0 0

22:01-23:00 0 0 0 0 0 0 0 0 0 0 0 0

23:01-24:00 0 0 0 0 0 0 0 0 0 0 0 0

Source: [ASHRAE, 2001]

2.2.4: Map and standards

Topographic map with scale 1:200000 were digitized and used for the spatial and temporal SWQ

distribution maps. Printed-paper maps were obtained from ALAGC.

The following two Mongolian National standards were used for the water sample collection and

SWQ assessment;

− MNS ISO 5667-6:2001 “Water quality. Sampling, 6th part. Guideline for sampling from

river and stream”

− MNS 4586:98 “Surface water quality. Permissible level of surface water variables”

Those standards were collected from CSM.



Table 11: Permissible level of surface water variables

Variable name Chemical formula Unit Permissible level

Hydrogen ion activity pH 6.5 – 8.5

Dissolved oxygen O2 mg l-1 6 and 4

Biochemical oxygen demand - 5 day BOD5 mg l-1 3

Chemical oxygen demand - Manganese COD-Mn mg l-1 10

Ammonium NH4-N mg l-1 0.5

Nitrite NO2-N mg l-1 0.02

Nitrate NO3-N mg l-1 9.0

Phosphorus PO4-P mg l-1 0.1

Chloride CI- mg l-1 300

Fluoride F mg l-1 1.5

Sulphate SO4-2 mg l-1 100

Manganese Mn mg l-1 0.1

Nickel Ni mg l-1 0.01

Copper Cu mg l-1 0.01

molybdenum Mo mg l-1 0.25

Cadmium Cd mg l-1 0.005

Cobalt Co mg l-1 0.01

Lead Pb mg l-1 0.01

Arsenic As mg l-1 0.01

Total Chromium Cr mg l-1 0.05

Chromium with hexavalence Cr+6 mg l-1 0.01

Zinc Zn mg l-1 0.01

Mercury Hg mg l-1 0.1

Hint: DO is 6 in warm session and 4 in cold session (water covered by ice)

Source: [CSM, 1998]

2.2.5: Pollution source

Chemical monthly data of discharge from the CWTP, the largest pollution point source, were

provided and supported by the NGIC project. The CWTP is responsible for the collection and

treatment of industrial and domestic sewage water in the Ulaanbaatar city. First mechanical

treatment filter installed in 1969. Overall treatment capacity is 230,000 m3 day-1. Recently,

150,000-160,000 m3 day-1 domestic and industrial sewage water is passing through a treatment

plant. In CWTP laboratory, 45 chemical and biological analyses can do. From those, 15 analysis

methods meet ISO standards.

Most recent problem is the electricity supply. If electrical power cut off during an hour, 4,500 m3

wastewater flows into the Tuul River without any treatment [CWTP, 2006]. For the dataset see

Appendix 6.



Chapter 3: Water quality assessment

3.1: Method for surface water quality assessment

In Mongolia, surface water quality is estimated by two different methods;

1. Water quality grade developed by Water sector of MNE (former name)

2. Water quality index developed by Erdenebayar,Ya., and Bulgan,T

In this research, second method was used to evaluate the surface water quality.

The water quality index is estimated using the following formulae:

Equation 1: Surface water quality index

n

Pl

C

W i

i

i

qi

Σ

=

Where:

Wqi water quality index

Pli permissible level of i-th variable

n number of variables

Ci concentration of i-th variable

The Mongolian National Standard MNS 4586-98, which was developed by CSM in 1998, determine

the maximum permissible levels of chemical variables in the surface water. Permissible levels (see

table 11 in chapter 2), DO, BOD5, COD-Mn, ammonium, nitrate, nitrite and other variables should be

included in the calculation of quality index [Davaa, et al., 2006].

After calculation of quality index, the water sample is classified using this system (Table 12):

Table 12: Assessment of surface water quality

Water quality Quality index

degree class

< 0.30 1 Very clean

0.31 – 0.89 2 Clean

0.90 – 2.49 3 Slightly polluted

2.50 – 3.99 4 Moderately polluted

4.00 – 5.99 5 Heavily polluted

6.00 < 6 Dirty

Source: [MNE, 2006]

Surface water usage depends on quality of the water. The MNE of Mongolia determines the definition

of the WQ classes and water usage of specific waters.



Table 13: Definition for classification of surface water quality in Mongolia

Classes Definition

1

Very clean

Sections of water bodies with pure, usually oxygen saturated, nutrient-poor water; low

bacteria content and directly use for drinking purpose, suitable for all kinds of water

usage.

2

Clean

Water bodies with small amount of inputs of organic or inorganic nutrients but without

or slightly oxygen depletion, low bacteria content and use for drinking and food

production purposes after disinfection and filtration, directly utilize for fishing factory.

3

Slightly

polluted

Section of water bodies with slight pollution, not a good oxygen supply, inputs of

organic or inorganic nutrients, some bacteria content and not suitable for drinking and

food production purposes, if no choice use it above mentioned purposes after treatment,

disinfection, filtration and can use directly for livestock, recreation, sport purposes.

4

Moderately

polluted

Water bodies with inputs of organic, oxygen consuming substances capable of producing

critical oxygen depletion; fish kills possible during short periods of oxygen deficiency;

declining numbers of macro-organisms; certain species tend to produce massive

populations and use for irrigation, industrial process after filtration.

5

Heavily

polluted

Sections of water bodies with heavy organic, usually low oxygen content; localised

deposits of anoxic sediment; filamentous sewage bacteria, occasional mass development

of a few micro-organisms, which are not sensitive to oxygen deficiency, periodic fish

kills occur, after filtration, use for some industrial process, which is not take part in

human.

6

Dirty

If value of water quality index exceeds 5th degree, it is belongs to this class. Sections of

water bodies with excessive pollution by organic, oxygen depleting sewage; processes of

putrefaction predominate; prolonged periods of very low oxygen concentrations or total

deoxygenating; mainly colonised by bacteria, no fish stocks; loss of biological life in the

presence of severe toxic inputs and can not use any purpose.

Source: [MNE, 1997b]

3.2: Basic geo-statistical analysis

The basic geo-statistical analysis was done by specific software “R” version 2.5.1, which developed in

2001 and recently updated in 27.06.2007. Chemical dataset of the Tuul River consists of totally 11

years data that were measured by CLEM between 1996-2006. Chemical variables, DO, BOD5, COD-

Mn, NH4+, NO2

-, NO3-, used to calculate water quality index.

3.2.1: Descriptive statistics

The Tuul River data frame contains two coordinates (here named E (UTM) and N (UTM)), two

categorical variable (name of sampling point) and ID. Four continuous variables, representing

water quality of the Tuul River expressed by index, time steps starting from 1996 until 2006 and

the distance measured from the upper reach by meter, water quality classified in 6 different

degrees. Totally, 1192 observations were done at 14 sampling points along the Tuul River and its

tributaries between 1996 and 2006.



Figure 9: Data frame of the Tuul River dataset

Table 14: Statistical summary of water quality index

Minimum 1st quartile Median Mean 3rd quartile Maximum

0.15 0.33 0.58 1.68 1.45 34.2

The minimum and maximum values of WQI are 0.15 and 34.2, respectively. The difference

between minimum and maximum value is 34.05.

Figure 10: Highest value of WQI

The highest value of WQI is 34.2 in row 852. This attribute value measured in December 2004, a

sample took at Tuul-Songino (down), which is located in 625304 East, 5300732 North and

distance from upper reach is 61017 m. In addition, almost all high values were measured at

sampling points in downstream, after the discharge from the CWTP enters the river.

According to the classification of surface water pollution in Mongolia, the first critical value in

river water is 0.3 and between 0.31 and 0.89 surface water concerns second level of water quality,

namely clean. Equal to or greater than 0.9 until 2.49, it belongs to third level and from 2.5 to 3.99

is fourth level. Fifth level has value between 4.0 and 5.99. Greater than or equal to 6.0, it concerns

sixth level and that water is strictly forbidden to be used for any purpose.

Number of observations with threshold values and its percentage in the entire dataset can be

estimate using the logical expressions in the R software.

Table 15: The quantity of samples with critical values

Threshold values Number of observations Percentage in total observations

<= 0.30 244 20.5

>= 0.31 948 79.5

>= 0.90 453 38.0

>= 2.50 167 14.0

>= 4.00 97 8.1

>= 6.00 74 6.2

Total numbers of observations, less than and equal to 0.3, are 244 and represent 20.5 percent of the

total of 1192 observations. The numbers of observations, which greater than and equal to 6.0, are

74 and get 6.2 percent in total observations. Cumulated observations 948 catch 79.5% that belongs

to observation with greater than and equal to 0.31.



3.2.2: Exploratory graphics

The R geo-statistical software package provides a rich environment for statistical visualization.

There are two graphics systems: the base system (in the graphics package, loaded by default when

R starts) and the Trellis system (implemented in R by the lattice package) [Rossiter, 2007]. The

subsequent figures visualized by the base system.

Figure 11: Histogram and box plot of water quality index

Histograms have one primary deficiency – their visual impression depends on the number of

categories selected for the plot. Comparisons of shape and similarity of histograms of the same

data depend on the choice of bar widths and centre [Helsel and Hirsch, 1992].

This histogram visualizes the frequency distribution of water quality index. The distribution is

strongly right-skewed and high values are very rare. The modal value is in the 0 to 2 range.

Another useful univariate plot is the box plot. Box plot which gives a rough idea of the shape of a

unimodal distribution [Rossiter, 2007]. Box plot in Figure 11 visualizes the quartiles distribution

of the water quality index. Most of attribute values are between 0 and 2.

Figure 12: Box plot for spatial variability of WQI

The second box plot shows WQ variability at

sampling points along the Tuul River. Most

dynamic one is a sampling point 11, namely

Tuul-Songino (down). At this point, WQ is

heavily influenced by treated water discharges

from the CWTP. That means, the variability of

WQI is directly related to human activity. Then,

it is naturally purified along the river in

downstream direction. First 10 sampling points

have less variability of WQI due to lesser

human impact (except some tourist camps and

towns) to river water quality.



3.3: Spatial water quality assessment

At a given river station water quality depends on many factors, including: (i) the proportion of surface

run-off and groundwater, (ii) reactions within the river system governed by internal processes, (iii) the

mixing of water from tributaries of different quality (in the case of heterogeneous river basins), and

(iv) inputs of pollutants [WHO, 1996].

Spatially, river water quality decreases along the river in downstream direction. Several point and

non-point contamination sources exist in study area. The point sources of pollution in the Tuul River

are improper treated wastewater from (i) Nalaih (1,400 m3 day-1), (ii) Niseh (400 m3 day-1), (iii)

CWTP (190,000 m3 day-1), (iv) Bio-industry (490 m3 day-1) and (v) Bio-Songino wastewater treatment

station (600 m3 day-1). The largest point source is CWTP, which is situated in western side of UB.

Daily mean discharge from that point is approximately 190,000 m3 day-1. The CWTP is responsible

for the collection and treatment of industrial and domestic sewage in the Ulaanbaatar city. Main non-

point source is the UB city that produces 800-1,000 tons of dry wastes per day [Orchlon, 1995].

Water pollution of the Tuul River basin is continuously increasing from upper to its lower reach.

Naturally, upstream of the Tuul River is running through mountainously area and there has high

velocity and turbulent. Hence, upper part of river has more oxidization potential, re-aeration and self-

purification. When it comes to the Ulaanbaatar city, natural conditions change from mountainous

region to valley. In the valley, velocity and turbulence of river decreases, then capability of

oxidization, re-aeration and self-purification also reduces as well. This is the natural factor of

possibility to store contaminant elements in the river water a longer time and distance [Altansukh,

2000].

Scatter plots were produced to illustrate the spatial distributions of the WQI.

Figure 13: Correlation between distance & WQI

Until around 60 km from upper reach, the water

quality is almost stable (comparing with high peak

value) which means did not reach to highest critical

value. Nevertheless, high peaks are starting from 11th

sampling point. Because of, that reach is joint part of

the Tuul River and discharge from the CWTP.

Based on above analysis, entire Tuul River dataset

separated into two sub-datasets, which are upstream

and downstream. Reason of this, the Tuul River

dataset consists of two different populations,

statistically. The upstream dataset contains chemical

analysis data of the Tuul River water quality from

upstream (sampling point number 1) until sampling

point number 10 which is located just upper reach of effluent from the CWTP. And, the downstream

dataset contains data from sampling point number 11 that is located lower reach of runoff from the

CWTP till last sampling point number 14 of this study.



Figure 14: Correlation between distance and WQI in the upstream section

0123456789

0 10000 20000 30000 40000 50000 60000

Distance, m

Ind

ex

In the upstream reach, there is little or no correlation between distance and water quality (nearly

stable). Fluctuation of water quality slightly changes along the river. Moreover, the quality index does

not reach the maximum critical value of 6 (ignoring some outliers). Two point pollution sources out of

five operate in upstream study area, namely Nalaih and Niseh WTSs. Total amount of effluents,

which are released from those two points, are approximately 1,800 m3 day-1. This amount of discharge

does not have a strong affect on the river (annual average flow 20 m3 s-1 > 0.021 m3 s-1). In addition,

distance between two point sources is around 54 km along the river. This is an adequate amount of

distance for river self-purification after the first effluents from Naliah WTS pour into the river water.

Figure 15: Water quality fluctuation along the upstream in 1996-2006

0

1

2

3

4

5

6

Ter

elj -

Ter

elj

Tuu

l -U

ubul

an

Tuu

l - N

alai

h

Tuu

l -B

ayan

zurh

Ulia

stai

- U

B

Tuu

l - Z

aisa

n

Tuu

l -S

onsg

olon

Tuu

l -S

ongi

no(u

pper

)

Name of sampling point

Inde

x

In the downstream reach, from a main pollutant source onwards, the quality index inversely correlates

to distance. The values in Figure 16 show that the index already exceeds the maximum critical value.

Because of, that is largest point source of contamination. Three point sources are situated in

downstream. Total volume of discharges from the CWTP, Bio-industry and Bio-Songino WTSs are

approximately 191,090 m3 day-1 and distance between point sources is around 2.5 km. This distance is

not sufficient for the self-purification process to take fully place, especially after large volumes of

effluent from CWTP pour into the river (annual average flow 20 m3 s-1 > 2.2 m3 s-1).



Figure 16: Correlation between distance and WQI in the downstream section

0

10

20

30

40

60000 65000 70000 75000 80000 85000 90000 95000

Distance, m

Ind

ex

When effluent from any source joins to the river, physical, biological and chemical processes such as

dilution, dispersion, advection and re-aeration occur in the river water, simultaneously. Those

processes are very beneficial in terms of water quality. Because river water has a self-purification

capacity, as a result of those processes. Self-purification is spatial and temporal process and depends

on many factors such as water flow velocity, river meandering, volume of fluid, chemical contents

and concentration, dilution, soil characteristics, diffusion, water turbulence, advection, temperature,

re-aeration and etc. Pollution of the Tuul River reduces along the downstream direction, but it is not

yet completely purified after a downstream distance of 35 km (after the main outfall).

3.4: Temporal water quality assessment

Water quality and pollution in surrounding area of the UB city was monitored in 14 stationary points

since 1980s. Ten sampling points along the Tuul River and 4 points in tributaries of the Tuul River, 1

in Terelj, 1 in Uliastai, 2 in Selbe, were chosen by CLEM in 1980th. In the last decade, fast

urbanization in combination with increasing number of tourist camps, agricultural and mining

activities have had a significant negative impact on the Tuul River’s quality and its associated

ecosystems. Consequently, the water became seriously polluted, lost its clarity and transparency and

its self-purification distance and time increased in year by year [Baasandorj and Davaa, 2006].

Figure 17: Correlation between time & WQI (1996-2006)

0

5

10

15

20

25

30

35

40

Jan-96 Jan-97 Jan-98 Jan-99 Jan-00 Jan-01 Jan-02 Jan-03 Jan-04 Jan-05 Jan-06

Time steps

Inde

x



Therefore, the general trend of water quality index and variability is increasing throughout the study

period. In the year 1999, the water quality was most stable and a year with high variability is 2005.

Reason of that is a new filtering system installed in the CWTP in 1999 by JICA support. Moreover,

this system is not renewed until now. In addition, amount of wastewater that flows through the

CWTP, is continuously increasing due to population growth of the capital city Ulaanbaatar and

industrialization.

Figure 18: Water quality fluctuation at Tuul-Uubulan SP during study period

0.34 0.32 0.29 0.32 0.30 0.330.28

0.350.32

0.43 0.41

0.00

0.10

0.20

0.30

0.40

0.50

1996 1998 2000 2002 2004 2006

Time step

Inde

x

Generally, the water quality did not snow large changes at selected sampling points in the upstream

reach during the study period. No large contamination sources of influence exist yet in the upper

reaches.

Figure 19: Water quality fluctuation at Tuul-Songino (upper) SP during study period

0.46

0.34

0.49

0.35 0.38 0.35

0.47

0.56 0.530.46

0.39

0.00

0.10

0.20

0.30

0.40

0.50

0.60

1996 1998 2000 2002 2004 2006

Time step

Inde

x

However, the water quality shows a slightly increment of general trend in upstream. In the

downstream section, water quality decreased during the study time and the general trend was an

increase in the WQI.

Figure 20: Water quality fluctuation at Tuul-Songino (down) SP during study time

3.014.33 3.93

1.44

2.904.12

7.77

5.926.74

7.458.60

0.00

2.00

4.00

6.00

8.00

10.00

1996 1998 2000 2002 2004 2006

Time step

Inde

x



The following time series of thematic WQ maps show the temporal changes of water quality along the

Tuul River in 3 selected years.

Figure 21: Water quality along the Tuul River in 1996

In the 1996, the Tuul River was not seriously polluted, yet. Classes of heavily polluted and dirty water

are not visible in the map.


In the 2002, the Tuul River starts getting seriously polluted. Classes of heavily polluted and dirty

water can be visualized in the map. Because of, rapid urbanization and increase of industrial activities,

including frequent spills from industries and insufficient operation of the CWTP.




In the 2006, the Tuul River starts

getting strongly polluted. Because

of, the efficiency of the CWTP

operation began to fail due to poor

maintenance, lack of spare parts,

outdated equipment, frequent

power shortages and plus above-

mentioned reasons.

3.4.1: Seasonal water quality assessment

In fact, Mongolia has four seasons. In this study, all months of the year were divided into two

periods, warm and cold. Cold period continues from November until end of March and average air

temperature is below zero and deposits snow. This natural phenomena negatively affects the

internal processes in the river and the interaction between the river and other natural components.

That means, self-purification process cannot takes place in this period. Because of, river discharge

reaches between 0 - 2 m3 s-1 and discharge from the CWTP is 2.2 m3 s-1, normally. Warm period

continues from April until end of October and mean air temperature is above zero, with rainfall.

Figure 24: Seasonal water quality in upstream reach

0.00

0.20

0.40

0.60

0.80

1.00

Tuul-Uubulan Tuul-Nalaih Tuul-Bayanzurh

Tuul-Zaisan Tuul-Sonsgolon

Tuul-Songino(upper)

Sampling point

Index

Overall mean in warm period Overall mean in cold period

In the upstream section, there is no big pollutant source (ignoring effluent from Nalaih WTS of

approximately 0.02 m3 s-1). That is why, the water quality index calculated with small values and

difference between indices in warm and cold period is little.

Figure 25: Seasonal water quality in downstream reach

0.00

2.00

4.00

6.00

8.00

10.00

12.00

Tuul-Songino (down) Tuul-Chicken farm Tuul-Khadanhyasaa Tuul-Altanbulag

Sampling point

Index

Overall mean in warm period Overall mean in cold period



In downstream section, several pollutant point sources exist and total discharge is 2.2 m3 s-1,

approximately. Water quality index calculated with high values in cold period and difference

between warm and cold period’s indices is high. Because of, self-purification process intensively

takes place in warm period and this process is almost interrupted during the cold period. That is

why, overall mean index of water quality calculated maximum value with 2.94 in warm period and

11.05 in cold period. See Table 16.

Table 16: Annual mean water quality index

Date Tuul-

Uubulan Tuul-Nalaih

Tuul-Bayanzurh

Tuul-Zaisan

Tuul-Sonsgolon

Tuul-Songino (upper)

Tuul-Songino (down)

Tuul-Chicken

farm

Tuul-Khadanhyasaa

Tuul-Altanbulag

Warm period Average-96 0.40 0.42 0.46 0.39 0.56 0.56 1.13 1.84 1.75 1.42 Average-97 0.35 0.32 0.41 0.35 0.40 0.37 1.56 1.87 1.65 1.44 Average-98 0.36 0.36 0.48 0.35 0.38 0.48 3.26 2.73 2.56 2.01 Average-99 0.36 0.40 0.40 0.36 0.59 0.36 0.91 1.36 1.50 1.36 Average-00 0.36 0.36 0.34 0.37 0.34 0.38 2.27 2.48 3.11 1.55 Average-01 0.34 0.38 0.41 0.38 0.37 0.36 2.36 2.11 1.98 1.37 Average-02 0.32 0.38 0.33 0.63 0.55 0.54 4.26 3.31 2.65 2.18 Average-03 0.41 0.47 0.45 0.44 0.46 0.60 4.55 6.52 1.33 2.69 Average-04 0.39 0.51 0.48 0.54 0.48 0.58 1.27 1.68 1.96 1.58 Average-05 0.37 0.50 0.39 0.39 0.34 0.39 3.69 3.62 2.43 1.70 Average-06 0.36 0.47 0.38 0.34 0.41 0.41 6.87 4.85 2.54 2.00

Overall mean 0.37 0.42 0.41 0.41 0.44 0.46 2.92 2.94 2.13 1.75 Cold period

Average-96 0.40 1.06 0.40 0.30 0.35 7.47 7.59 2.00 1.26 Average-97 0.27 0.25 0.24 0.51 2.53 0.30 11.19 8.69 3.40 1.58 Average-98 0.24 0.29 0.27 0.30 0.30 0.26 3.65 6.09 1.82 0.77 Average-99 0.27 0.32 0.28 0.26 0.28 3.16 3.65 3.02 1.89 Average-00 0.28 3.44 0.52 7.08 3.53 1.51 0.68 Average-01 0.37 0.50 0.32 0.43 0.34 0.26 8.56 4.19 1.92 1.42 Average-02 0.28 0.50 0.35 0.29 15.20 10.52 4.11 3.17 Average-03 3.42 0.85 0.32 0.39 12.10 5.52 2.57 Average-04 0.55 0.39 0.25 0.17 0.17 17.78 12.62 5.21 3.25 Average-05 2.05 0.73 0.48 1.05 1.15 0.87 17.76 13.40 9.36 3.01 Average-06 0.65 0.62 0.35 0.23 0.21 0.23 17.57 13.07 5.59 20.80

Overall mean 0.80 0.81 0.33 0.42 0.65 0.36 11.05 8.08 3.68 3.78

3.5: Chapter conclusions

The following conclusion can be drawn from this chapter:

− The highest value of the WQI observed was 34.2. This attribute value was measured in

December 2004, in a water sample from Tuul-Songino (down).

− Statistically, the entire dataset consisted by two different populations, natural water and

natural water mixed with discharge from the CWTP.

− Until the main pollutant source outfall enters the Tuul River, there is no significant

correlation between distance and water quality. However, from the main pollutant source in

downstream, the water quality inversely correlates to distance.

− The general trend of water quality index and variability increases throughout the time steps.

In the year 1999, the water quality was most stable and a year with high variability is 2005.



− Pollution of the Tuul River reduces as a function of distance in the downstream section, but

completely self-purification was not observed even at the last sampling point of this study (35

km downstream major outfall).

− The natural phenomenon of cold period (t < 00C), negatively affects the self-purification

process of the Tuul river. Reason is also that the discharge from pollutant sources is in this

period greater than the natural river discharge.

− The CWTP remains the biggest pollution source on this section of the Tuul River.

3.6: Limitation in WQA

In this section, the correlation between water quality and quantity could not be studied in detail. One

reason is that the Tuul River hydraulic parameters measurement and quality analysis are implemented

by two different organizations. Moreover, measurement dates and locations of both organizations are

not coinciding. For example, daily discharge measurement are done in three different stations, which

belong to WS, NAMHEM, situated in study area. Water quality analysis, implemented by CLEM

once a month, is done in different locations from the hydraulic measurements. Furthermore, quality

analysis date is not written in the chemical dataset that provided by CLEM.

However, monthly mean discharge value and monthly WQI at Tuul-Zaisan sampling point were

plotted in the following graph.

Figure 26: Correlation between river monthly mean discharge and monthly WQI

Correlation between river discharge and WQ(Tuul-Zaisan)

0.1

0.3

0.5

0.7

0.9

1.1

Ap

r-9

6

Ap

r-9

7

Ap

r-9

8

Ap

r-9

9

Ap

r-0

0

Ap

r-0

1

Ap

r-0

2

Ap

r-0

3

Date

WQ

I

0.00

20.00

40.00

60.00

80.00

100.00

Q, m

3 s

-1

WQI Q

Obviously, water quality can be related to quantity. In general, when river discharge increases then

WQI decreases (only case of effluent from pollution source is constant). However, we can observe

hysteresis effects and large variations and deviations can be observed.



Chapter 4: River flow model

4.1: General introduction of flow model

The Duflow is a linked flow and water quality modelling package, which can be used to simulate one

dimensional flow and water quality in a network of open water courses [DMS, 2004b].

The Duflow is a numerical model. To model the water system network, the real water system has to

be sub-divided into sections and nodal points. Sections reflect river reaches of uniform characteristics

and nodal points are used to connect the sections. For each section characteristics, dimensions, like

length and cross sectional profile, have to be defined. Manning or Chezy coefficients can be used to

account for section roughness [Makkinga, et al., 1998]. Within the network, several types of

structures can be defined (culverts, overflows, underflows, siphons and pumps). The flow model also

includes a simple rainfall-runoff module. At each nodal point, a catchment area can be defined and

using a runoff coefficient, rain losses can be taken into account. Furthermore, at each nodal point

additional flows can be defined to account for discharges and tributaries not included in the network.

At the system boundaries additional flows, discharge curve or water levels can be used as boundary

conditions. All boundary conditions and discharges can be entered as constants or as time functions

[Makkinga, et al., 1998].

4.2: Background theory of flow simulation

The following equations mainly used to simulation and evaluation of a flow model.

4.2.1: Saint-Venant equation

DUFLOW simulation is based on the one-dimensional partial differential equation (Saint-Venant)

that describes non-stationary flow in open channels (Abbott, 1979; Dronkers, 1964).

Several assumptions consider for St. Venant equations:

− Flow is one-dimensional

− Hydrostatic pressure prevails and vertical accelerations are negligible

− Streamline curvature is small

− Bottom slope of the channel is small

− Manning equation is used to describe resistance effects

− The fluid is incompressible [Merwade, 2005]

These equations, which are the mathematical translation of the laws of conservation of mass and of

momentum:

Equation 2: Mass equation

0x

Q

t

B =∂∂+

∂∂

Equation 3: Momentum equation

( ) ( )φ−Φγ=+∂α∂+

∂∂+

∂∂

coswaARC

QQg

x

Qv

x

HgA

t

Q 22



Equation 4: Discharge equation A*vQ =

Where:

t time [s]

x distance as measured along the channel axis [m]

H(x, t) water level with respect to reference level [m]

v(x, t) mean velocity (averaged over the cross-sectional area) [m s-1]

Q(x, t) discharge at location x and at time t [m3 s-1]

R(x, H) hydraulic radius of cross-section [m]

a(x, H) cross-sectional flow width [m]

A(x, H) cross-sectional flow area [m2]

b(x, H) cross-sectional storage width [m]

B(x, H) cross-sectional storage area [m2]

g acceleration due to gravity [m s-2]

C(x, H) coefficient of De Chézy [m1/2 s-1]

w(t) wind velocity [m s-1]

Φ(t) wind direction [degrees]

φ(x) direction of channel axis, measured clockwise from the north [degrees]

γ(x) wind conversion coefficient [-]

α correction factor for non-uniformity of the velocity distribution in the advection

term, defined as:

Equation 5: Correction factor for non-uniformity of velocity distribution

( )∫=α dydzz,yvQ

A 2

2

A cross-section [m2]

The mass equation states, if the water level changes at some location this will be the net result of

local inflow minus outflow. The momentum equation expresses that the net change of momentum

is the result of interior and exterior forces like friction, wind and gravity. For the derivation of

these equations, it has been assumed that the fluid is well mixed and hence the density may be

considered constant [DMS, 2004b].

4.2.2: Manning equation

The Manning equation is the most commonly used equation to analyze open channel flows. It is a

semi-empirical equation for simulating water flows in channels and culverts where the water is

open to the atmosphere, i.e. not flowing under pressure, and was first presented in 1889 by Robert

Manning. The channel can be any shape - circular, rectangular, triangular, etc. The units in the

Manning equation appear to be inconsistent; however, the value k has hidden units in it to make

the equation consistent [Edwards, 1998].

Equation 6: Hydraulic radius

P

AR =



Equation 7: Slope of riverbed

L

hS f∆

=

Equation 8: River velocity

2

1

3

2

SRn

kV =

Where:

V velocity, m s-1

A cross-sectional area, m2

n Manning coefficient

P wetted perimeter, m

R hydraulic radius, m

S slope

hf difference between bottom of channels, m

L length of channel, m

k 1.0 for unit conversion

Slope of selected section of the Tuul River is approximately 0.00198 or 0.2%. Reason of calculate

the slope, if slope exceeds 2% then DMS cannot handle it.

4.2.3: Froude number

The Froude number, Fr, is a dimensionless value that describes different flow regimes of open

channel flow. The Froude number is a ratio of inertial and gravitational forces.

− Gravity (numerator) - moves water downhill

− Inertia (denominator) - reflects its willingness to do so [White, 1998].

Equation 9: Froude number

gD

VFr =

Where:

D hydraulic depth, m

g gravity, m s-1

A cross-sectional area, m2

b flow width, m

Equation 10: Hydraulic depth

b

AD =

When:

Fr = 1, critical flow,

Fr > 1, supercritical flow (fast rapid flow),

Fr < 1, subcritical flow (slow / tranquil flow)



The Froude number is a measurement of bulk flow characteristics such as waves, bed forms, and

flow, depth interactions at a cross section or between boulders. The denominator represents the

speed of a small wave on the water surface relative to the speed of the water, called wave celerity.

At critical flow celerity equals flow velocity. Any disturbance to the surface will remain stationary.

In subcritical flow the flow is controlled from a downstream point and information is transmitted

upstream. This condition leads to backwater effects. Supercritical flow is controlled upstream and

disturbances are transmitted downstream [Furniss, et al., 2006].

At Ulaanbaatar station, Fr is equal to 0.74, which means subcritical flow.

74.0m63.0*s/m81.9

s/m85.1

gD

VFr

2===

At Altanbulag station, Fr is equal to 0.45, which means subcritical flow.

45.0m67.0*s/m81.9

s/m16.1

gD

VFr

2===

Maximum velocity and associated width, hydraulic area in study period are used to calculate

Froude number at specific location. Reason of Fr calculation, if flow is supercritical then DMS

cannot simulate the river flow.

4.2.4: Nash-Sutcliffe coefficient

The Nash-Sutcliffe model efficiency coefficient is used to assess the predictive power of

hydrological models. It is defined as:

Equation 11: Nash-Sutcliffe coefficient

( )( )∑

∑

=

=

−

−−=

T

1t

2

oto

T

1t

2tm

to

QQ

QQ1E

Where:

Qo observed discharge

Qm modelled discharge

Qt discharge at time t

Nash-Sutcliffe efficiencies can range from -∞ to 1. An efficiency of 1 (E = 1) corresponds to a

perfect match of modelled discharge to the observed data. An efficiency of 0 (E = 0) indicates that

the model predictions are as accurate as the mean of the observed data, whereas an efficiency less

than zero (-∞<E<0) occurs when the observed mean is a better predictor than the model.

Essentially, the closer the model efficiency is to 1, the more accurate the model is [Nash and

Sutcliffe, 1970].



4.2.5: Statistical methods ME, MAE, RMSE

Mean error measures the average between observed and calculated variables. Both negative and

positive differences are incorporated in the calculation; they may cancel out the error. As such

small error may not indicate a good calibration [Rientjes, 2006].

Equation 12: Mean error

( )∑ −−=

n

1ico

n

1ME

Where:

n number of values

o observed values

c calculated values

The mean absolute error measures the average magnitude of the errors in a set of forecasts, without

considering their direction. It measures accuracy for continuous variables. The MAE is the average

over the verification sample of the absolute values of the differences between forecast and the

corresponding observation. The MAE is a linear score, which means that all the individual

differences are weighted equally in the average.

Equation 13: Mean absolute error

∑ −−=

n

1ico

n

1MAE

The root mean squared error measures the difference between forecast and corresponding observed

values are each squared and then averaged over the sample. Finally, the square root of the average

is taken. Since the errors are squared before they are averaged, the RMSE gives a relatively high

weight to large errors. This means the RMSE is most useful when large errors are particularly

undesirable.

Equation 14: Root means squared error

( )5.0

n

1i

2ico

n

1RMSE

−= ∑ −

4.3: Flow model setup

An open channel flow model of the Tuul River is one-dimensional, distributed2, mathematical3,

conceptual4 and non-steady state5 hydraulic model.

The following flow chart is shown general sequence of flow modelling. All input data can divided

into two different types:

− Field data measured in field

− Non-field data assumed, constant and adjustable values

2 Model domains are discretised in space by use of uniform or non-uniform grid elements. 3 Partial differential equations are used in model. 4 Mathematical relations are applied to simulate the observed real world behavior. 5 A time variable is calculated for each calculation time step.



Figure 27: Flowchart of a flow model

The flow of the Tuul river system has been modelled, calibrated and validated using some existing

data between 2005 and 2006. The following datasets are used to simulate a flow model.

1. Geographical locations of hydraulic stations, WTP and true value of distance between

stations, discharge points

2. Daily discharge data at Tuul-Ulaanbaatar station in 2005-2006

3. Daily discharge data at Tuul-Altanbulag station in 2005-2006

4. Monthly lumped discharge data of CWTP

5. Monthly lumped discharge data of Bio WTP

6. Hydraulic data at Tuul-Ulaanbaatar station

7. Hydraulic data at Tuul-Altanbulag station

8. Simplified cross-section of the Tuul River under Zaisan bridge

9. Simplified cross-section of the Tuul River under Altanbulag bridge

4.3.1: Network development

Network of the Tuul River model consists of the following objects:

− Two pink nodes are representing begin and end boundaries

− Sections

− Cross-sections (simplified)

− Two discharge points are indicating CWTP and Bio WTP

− Most lower node is demonstrating Tuul-Altanbulag station, that was used for calibration



Figure 28: A flow model network

A flow model network contains three layers; green areas are representing settlement areas, blue

line is the Tuul River and red squire is located in real position of CWTP. A begin node, right side,

is situated in 643389 E, 5305700 N and an end node, left side, is placed in 587593 E, 5301654 N.

Position of nodes (except for an end node), discharge points, length of sections (sum of length is

approximately 54 km), are all true values. However, position of an end node and length of last

section are false value that is assumption. A flow model setup is shown in Appendix 7.

Figure 29: Simplified cross-section of the Tuul River under Zaisan Bridge

Type of simplified upper cross section is trapezoid and floor level is 1285.25 m, surface level is

1292.6 m. Floor width of river channel is 48.6 m, maximum depth is 1.5 m and approximate area is

95.4 m2.

This simplified cross-section based on hydraulic measurement at Tuul-Ulaanbaatar station, which

implemented by WS, NAMHEM. Minimum and maximum depth, width of river is used to develop

simple cross-sections. For example, minimum depth of river was 0.38 m and maximum was 0.98 m.

Width of river at that time was 55.9 m and 67.7 m, correspondingly. The way of developing trapezoid

cross-section is placing minimum and maximum depth, width and draw straight lines to connect two



depths. Finally, slope angle calculated using a linear equation 15 and used this simplified cross-

section in DMS model. In this case, slope angle is 5.710.

Equation 15: Slope angle equation

−−

=∠12

12

xx

yyarctanS

Figure 30: Simplified cross-section of the Tuul River under Altanbulag Bridge

Type of middle cross section is trapezoid and floor level is 1178.3 m, surface level is 1182 m.

Floor width is 26.6 m, maximum depth is 1.5 m, slope angle is 5.40 and approximate area is

63.9 m2. Cross-sectional data interpolated along the river and slope of riverbed is 0.00198.

Note: If study area located higher than 1000 meter above sea level, then this software cannot simulate it.

This is one weakness (bug) of Duflow modelling system, which we discovered during our research.

This limitation not reported in the user guide of DMS. Then we decided to deduct 1000 from level

value and used it into model. For example, our actual value of floor level is 1285.25 m, but we

used 285.25 m instead of real one. Actually, this change has no effect in model. Reason of that is

difference of level values is used to calculate slope of riverbed.

To turn a network schematization in a flow model the following three steps has to be taken:

− Define initial condition

− Define boundary condition

− Configure the calculation [DMS, 2004b]

4.3.2: Initial condition of flow model

To start the computations, initial values for all state variables are required. These initial values

supplied for each node [DMS, 2004a]. Values, which used in initial condition, obtained from field

measured data and former computations.



Table 17: Initial condition in flow model

Model structure Sampling point Discharge in m3 s-1 Level in m

Section 1 – begin Tuul-Zaisan 0.198 286.44

Section 1 – end Tuul-Sonsgolon 0.172 265.54

Section 2 – begin Tuul-Sonsgolon 0.172 265.54

Section 2 – end Tuul-Songino (upper) 0.155 249.7

Section 3 – begin Tuul-Songino (upper) 0.155 249.7

Discharge point 1 – begin CWTP 0.155 249.7

Discharge point 1 – end CWTP 0.155 249.7

Section 3 – end Tuul-Songino (down) 2.039 247.82

Section 4 – begin Tuul-Songino (down) 2.039 247.82

Discharge point 2 – begin Bio WTP 2.036 245.83

Discharge point 2 – end Bio WTP 2.036 245.83

Section 4 – end Tuul-Chicken farm 2.025 228.02

Section 5 – begin Tuul-Chicken farm 2.025 228.02

Section 5 – end Tuul-Khadanhyasaa 2 212.18

Section 6 – begin Tuul-Khadanhyasaa 2 212.18

Section 6 – end Tuul-Altanbulag 1.863 184.17

Section 7 – begin Tuul-Altanbulag 1.863 184.17

Section 7 – end Extension part 1.38 174.8

4.3.3: Boundary condition of flow model

In the flow model, boundary conditions should be defined at beginning and end nodes. Discharge

of the Tuul River at Tuul-Ulaanbaatar station, which measured in 2005 and 2006 by WS,

NAMHEM, is a boundary condition of beginning node. The type of boundary condition is an

equidistant (daily) starts from 01 April 2005 until 30 November 2006. Input data covers only warm

session. Reason of that is river freezes during cold session and consequence of frozen river is no or

less than 0.1 m3 s-1 discharge.

Figure 31: Discharge of the Tuul River under Zaisan Bridge in 2005 and 2006

0.000

20.000

40.000

60.000

80.000

100.000

120.000

140.000

160.000

3/1

/05

4/1

/05

5/1

/05

6/1

/05

7/1

/05

8/1

/05

9/1

/05

10/1

/05

11/1

/05

12/1

/05

1/1

/06

2/1

/06

3/1

/06

4/1

/06

5/1

/06

6/1

/06

7/1

/06

8/1

/06

9/1

/06

10/1

/06

11/1

/06

12/1

/06

Date

Q in

m3/s



An artificial water discharge outflow was selected as a boundary condition of the end node and

type is constant in 10 days. Time step and covered period are same with previous.

Figure 32: Artificial discharge at end node

-35.00

-30.00

-25.00

-20.00

-15.00

-10.00

-5.00

0.00

3/1

/20

05

4/1

/20

05

5/1

/20

05

6/1

/20

05

7/1

/20

05

8/1

/20

05

9/1

/20

05

10/

1/2

005

11/

1/2

005

12/

1/2

005

1/1

/20

06

2/1

/20

06

3/1

/20

06

4/1

/20

06

5/1

/20

06

6/1

/20

06

7/1

/20

06

8/1

/20

06

9/1

/20

06

10/

1/2

006

11/

1/2

006

12/

1/2

006

Date

Q in

m3/

s

4.3.4: Calculation setting

Before running model, calculation setting should be set and computation starts same with input

data. In this prototype model, the following calculation setting was defined:

− Starting from 01 March, 2005 till 1 December, 2006

− Time step is 1 minute and output is daily

− Theta equals to 0.9

− Distance between calculation points is 200 m

− Advection term is “Total”

− Resistance formula is Manning

Data in March was used to stabilize the model calculation. Output of model starts from 1st of April,

2005.

4.4: Calibration

Actual discharge data at Tuul-Altanbulag station were used to calibrate the flow model. Discharge

data, which used in calibration, measured by WS, NAMHEM between 1st of April and 30th of

November 2005. Rest of months are belongs to cold session. Hydraulic measurements do not take

place in the cold session.



Figure 33: Correlation between calculated and observed discharge in calibration year

R2 = 0.8268

0.00

10.00

20.00

30.00

40.00

50.00

60.00

70.00

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00

Observed Q in m3/s

Cal

cula

ted

Q in

m3/

s

For the calibration year, correlation between calculated and observed values was evaluated by

different statistical approaches and the following results calculated in 2005 at Tuul-Altanbulag

station:

− ME -1.246

− MAE 2.539

− RMSE 4.162

− R2 0.827

− E 0.763

Figure 34: Observed and modelled Q in calibration year

0.00

10.00

20.00

30.00

40.00

50.00

60.00

70.00

4/1/2005 5/1/2005 6/1/2005 7/1/2005 8/1/2005 9/1/2005 10/1/2005 11/1/2005

Date

Q in

m3/s

Observed Modelled



4.4.1: Sensitivity analysis

The main purpose of a sensitivity analysis is to quantify how sensitive the model is to certain

changes in the input data. A sensitivity analysis basically is a procedure to quantify on the

uncertainty of the calibrated model [Rientjes, 2006].

Sensitivity analyse of the flow model was done by using different Manning values. Results of

analysis evaluated in dissimilar statistical methods such as MAE, RMSE, R2, Nash-Sutcliffe model

efficiency coefficient and ME.

Figure 35: Sensitivity analysis

0.000

10.000

20.000

30.000

40.000

50.000

60.000

70.000

4/1/2005 5/1/2005 6/1/2005 7/1/2005 8/1/2005 9/1/2005 10/1/2005 11/1/2005

Date

Q in

m3/s

Observed Modelled, n=0.04 Modelled, n=0.15

Manning values are less sensitive. However, some small changes were presented during analysis.

Figure 35 shown graphical representation of comparison observed and modelled values with

selected minimum and maximum Manning values.

Table 18: Statistical evaluation of sensitivity analysis

Different Manning values Statistical evaluation n=0.03 n=0.04 n=0.06 n=0.1 n=0.15 R2 n.a 0.821 0.828 0.837 0.841 ME n.a -1.218 -1.218 -1.218 -1.222 MAE n.a 2.534 2.520 2.501 2.503 RMSE n.a 4.192 4.128 4.040 4.035 E n.a 0.760 0.767 0.777 0.778

Totally, five different Manning values were selected for sensitivity analysis and evaluated by

different statistical methods. Based on graph and statistical assessment, the best result given by

n=0.15. Because of, R2 value is highest and RMSE is lowest one. Therefore, coefficient of model

efficiency is 0.778.

However, n=0.06 was applied in a flow model. Reason of that, characteristics of river channel is

coincide with 1f in table for the Manning values (see Appendix 8).



4.5: Validation

Actual discharge data at Altanbulag station was used to validate flow model. Discharge data, which

used in validation, measured by WS, NAMHEM between 1st of April and 30th of November 2006.

Figure 36: Correlation between calculated and observed Q data in validation year

R2 = 0.885

0.00

10.00

20.00

30.00

40.00

50.00

0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00

Observed Q in m3/s

Cal

cula

ted

Q in

m3/

s

In validation year, correlation between calculated and observed values was evaluated by different

statistical approaches and the following results calculated in 2006 at Tuul-Altanbulag station:

− ME 1.347

− MAE 2.677

− RMSE 4.781

− R2 0.885

− E 0.830

Figure 37: Observed and modelled Q in validation year

0.00

10.00

20.00

30.00

40.00

50.00

60.00

70.00

4/1/2006 5/1/2006 6/1/2006 7/1/2006 8/1/2006 9/1/2006 10/1/2006 11/1/2006

Date

Q in

m3

/s

Observed Modelled



4.6: Chapter conclusion

From a flow model simulation, the following conclusions can be drawn:

1. DMS is suitable for simulating the selected section of the Tuul River in Mongolia. Because of

the average slope of the riverbed is less than 1% (0.2%), yielding mostly sub-critical flow

regimes (i.e. Fr = 0.74), water levels and flows could be simulated with good accuracy in the

river reaches upstream and downstream of the Ulaanbaatar city.

2. Sensitivity analysis showed that the flow model is not very sensitivity to Manning values.

Difference between selected minimum and maximum values is R2=0.02 and E=0.018.

3. If real water level and river bed elevation values are greater than one thousand (m.a.s.l.), then

DMS shows an error message. Maybe, this is a bug of this software. If observed real data

values are used, they have to be rescaled to (<1000 m) values. It has minor effect to the

quality model in terms of re-aeration.

4.7: Limitation in a flow modelling

In the flow model simulation, interpretation of actual water level data in both stations, especially at

Tuul-Altanbulag station, was sometimes very confusing. According to some sources, artificial surface

level were used to measure the water levels. For example, we received values greater than 5 meter,

typed in water level data at Tuul-Altanbulag station. In reality and proven by field inspection, those

values were false. Water level measurements and artificial surface level data could therefore not be

used in the flow model calibration and validation experiment.



Chapter 5: Water quality model

5.1: Introduction of quality model

The DMS water quality model is usable for simulating the transportation of substances in free surface

quality. One can define its own substances and determine it is in or decline in concentration. There is

a standard parameter for natural, algorithmic decay depending on actual concentration. In DMS water

quality a distinction is made between substances, which are transported with the quality of the water,

for instance dissolved substances, and bottom materials that are not transported. This distinction

offers the user the facility to, for instance, study the interaction between the bottom materials and the

dissolved substances in the water above [DMS, 2007].

DMS is a flexible package, as the user is able to define his own set of process descriptions. This

means that except for the transport of pollutants within the network, the user can provide all other

physical, chemical and biological interactions between the state variables. The input for the

simulation of the water quality depends on the defined state variables and processes. Typically,

parameters like process rates, equilibrium constants and other model constants have to be entered.

Finally, external variables can be entered as constraints or as time functions (i.e. temperature and

irradiance) [Makkinga, et al., 1998].

Water quality modelling can be a valuable tool for water management since it can simulate the

potential response of the aquatic system to such changes as the addition of organic pollution, the

building of small hydro-electric power plants, the increase in nutrient levels or water abstraction rates

and changes in sewage treatment operations (such as the addition of tertiary treatment). Most existing

river models are for oxygen balance and are based on BOD measurements. Krenkel and Novotny

(1980) have listed the categories of variables required for oxygen balance modelling as:

− hydrological variables (e.g. river discharge),

− hydraulic variables (velocity, geometry of river bed, turbulence, etc.),

− oxygen sinks (e.g. benthic oxygen demand, nitrification of ammonia),

− oxygen sources (e.g. re-aeration, atmospheric exchange, primary production), and

− temperature

Downstream of sewage effluent discharges from treatment plants using biological, secondary

processes, bacterial activity may also need to be incorporated into the models [Krenkel and Novothy,

1980] and [WHO, 1996].

In system analysis, the processes affecting the oxygen concentration at a certain time at an exact point

in a water body can be schematized as:

− advection and dispersive mechanisms of oxygen transport in the water body

− exchange with the atmosphere (re-aeration)

− oxygen consumption for oxidation of organic matter by micro-organisms;

− biochemical oxygen demand and nitrification

− Oxygen demand of bottom sediment and benthic micro-organisms, SOD

− Primary production and respiration of photosynthetic i.e., the oxygen production and demand

of algae and water plants [Mannaerts, 2007]



Concentrations of dissolved oxygen in unpolluted waters range usually between 7 mg l-1 in warmer

and 14 mg l-1 in colder waters as they are a strong function of water temperature. Variations in DO

can occur seasonally, or even over 24 hour periods, in relation to temperature and biological activity

(i.e. photosynthesis and respiration). Biological respiration, including that related to decomposition

processes, reduces DO concentrations. In still waters, pockets of high and low concentrations of

dissolved oxygen can occur depending on the rates of biological processes. Waste discharges high in

organic matter and nutrients can lead to decreases in DO concentrations as a result of the increased

microbial activity (respiration) occurring during the degradation of the organic matter [WHO, 1996].

Sources of DO:

− Re-aeration from the atmosphere

− Photosynthetic oxygen production

− DO in tributaries or effluents

Sinks of DO:

− Oxidation of carbonaceous waste material

− Oxidation of nitrogenous waste material

− Oxygen demanding sediments of water body

− Use of oxygen for respiration by water plants [Thomann and Mueller, 1987]

5.2: Background theory of quality modelling

The mass transport equation, Peclet number, formulas for the transport of pollutants within the

network and physical, chemical, biological interactions between the state variables are part of the

mathematical background of quality modelling.

5.2.1: Mass transport equation

The quality part of the DUFLOW package is based upon the one dimensional transport equation.

This partial differential equation describes the concentration of a constituent in a one-dimensional

system as function of time and place.

Equation 16: Momentum balance equation

( )P

x

CAD

xx

)QC(

t

BC +

∂∂

∂∂+

∂∂−=

∂∂

Where:

C constituent concentration [g/m3]

Q quality [m3/s]

A cross-sectional quality area [m2]

D dispersion coefficient [m2/s]

B cross-sectional storage area [m2]

x x coordinate [m]

t time [s]

P production of the constituent per unit length of the section [g/m.s]

The production term of the equation includes all physical, chemical and biological processes to

which a specific constituent is subject. In order to apply this method equation 16 is rewritten as:



Equation 17: Mass balance equation

( )0P

t

BC

x

S =−∂

∂+∂∂

In which, S is the transport (quality of the constituent passing a cross-section per unit of time):

Equation 18: Equation for the constituent transport by advection and dispersion

x

CADQCS

∂∂−=

Equation 18 describes the transport by advection and dispersion. Equation 17 is the mathematical

formulation of the mass conservation law, which states that the accumulation at a certain location x

is equal to the net production rate minus the transport gradient [DMS, 2004a].

The dispersion coefficient either can be supplied by the user or can be calculated from the

characteristics of the quality. The empirical equation according to Fisher (1979) is used to

calculate the quality dependent part. In order to prevent the dispersion coefficient to become 0 at

low quality velocities a constant term is added that reflects background dispersion. The quality

independent part is in particular important in stagnant systems, where it represents the wind

induced mixing. The following equation is used:

Equation 19: Background dispersion coefficient

( ) 0

22s

x Du*Z

W*ut,xD +α=

Where:

αx a proportionality constant [-]

W quality width [m]

us average quality over the cross-sectional area [m/s]

Z water depth [m]

u shear stress velocity [m/s]

D0 background dispersion coefficient [m2/s]

The shear stress u can be written as:

Equation 20: Shear stress equation

C

guu s=

With:

C coefficient of De Chézy [m1/2/s]

g acceleration due to gravity [m2/s]

As all the characteristics of the quality are known or calculated in the quality part, above-

mentioned equations easily can be used to calculate the dispersion coefficient. Only αx and D0

have to be supplied by the user [DMS, 2004a].



5.2.2: Peclet number

In fluid dynamics, Peclet number is a dimensionless number relating the rate of advection of a flow

to its rate of diffusion. It is equivalent to the product of the Reynolds number with the Schmidt

number in the case of mass diffusion. For mass diffusion, it is defined as:

Equation 21: Peclet number

D

LVPe=

Where;

L characteristic length

V velocity

D mass diffusivity [Patankar, 1980]

5.2.3: DO balance equations

The amount of oxygen in the water is function of re-aeration, photosynthesis process, sediment

oxygen demand, oxygen consumption for biochemical process and nitrification.

Figure 38: Schematization of DO balance model

The following equations are used to DO model and initially developed by University of

Wageningen.

The DO balance in the river can be written as follows:

Equation 22: Dissolved oxygen balance

( ) ( ) ( ) ( ) ( ) ( )dt

Nitrifd

dt

BODoxd

dt

SedDOd

dt

PDOd

dt

OSd*

dt

KAd

dt

dDO ++++=

Where:

DO dissolved oxygen concentration in the water in mg l-1

KA re-aeration rate in day

OS oxygen saturation in mg l-1

PDO primary production in g m-2 day--1

SedDO oxygen consumption for sediment in g m-2 day-1

BODox oxidation of BOD in g m-2 day-1

Nitrif oxygen consumption for nitrification in g m-2 day-1



Re-aeration defined as:

− Process of oxygen absorption from the atmosphere

− Assumed a first order process; absorption rate is proportional to deficit

A basic and main theory of gas transfer, which commonly used in flowing waters such as river, is a

surface renewal model. Danckwerts (1951) modified Higbie’s (1935) the penetration theory by

assuming that the fluid elements reach and leave the interface randomly. The direction and

magnitude of the mass transfer depends partially on the difference between the saturation value

and the actual value in the water. If the water is unsaturated (o < os), then transfer will be positive

(a gain) as oxygen moves from the atmosphere into the water to try to bring the water back to the

equilibrium state of saturation. Conversely, if the water is supersaturated (o > os), then transfer

will be negative (a loss) as oxygen is purged from the system [Chapra, 1997].

The following series equations used for re-aeration rate:

Equation 23: Temperature dependent oxygen mass-transfer velocity

)20T(TKL*dt

20KL

dt

KLT −=

Equation 24: Re-aeration rate

Z

KLT

dt

KA =

Where:

TKL temperature coefficient of oxygen mass-transfer (1.024)

KLmin minimum oxygen mass-transfer rate in flowing system in m d-1 (0.1)

KL20 oxygen mass-transfer rate in the water laminar layer in m d-1

KLT temperature dependent O2 mass-transfer velocity in the water laminar layer in m d-1

KA re-aeration rate in day

When:

If KL20 < KLmin, then KL20 = KLmin

Oxygen saturation can be solved by: The saturation concentration of oxygen in natural water is on the order of 10 mg l-1. In general,

several environmental factors can affect this value. From the perspective of water quality

modelling, the most important of these are:

− Water temperature

− Salinity

− Partial pressure variations due to elevation

Several empirical derived equations have been developed to predict how these factors influence

saturation [Chapra, 1997]. However, only water temperature taken into account and the following

equation can be used to calculate the dependence of oxygen saturation on water temperature (0C):

Equation 25: Oxygen saturation

( ) ( ) ( ) ( ) ( ) ( ) ( )dt

Td*

dt

Td*

dt

Td*000077774.0

dt

Td*

dt

Td*007991.0

dt

Td*41022.0652.14

dt

OSd −+−=

With:

T water temperature in 0C



Primary production written as below: Photosynthesis creates oxygen and respiration depletes oxygen, the plants will have an impact on

the oxygen resources. Because photosynthesis is light dependent, this effect can have both

seasonal and diurnal manifestations. In most water quality models, the rate of photosynthesis is

assumed to be directly proportional to light energy. Oxygen level could be supersaturated during

the afternoon and severely depleted just before dawn [Chapra, 1997]. Primary production by

photosynthetic systems is functions of solar radiation, algal biomass and can be presented by:

Equation 26: Primary production

( )A*IO*

dt

PDOd β=

Where:

β oxygen production constant in (gO2 day-1. mg Chl-1) (W.m-2)-1

IO light intensity in W m-2

A chlorophyll concentration in ug l-1

Oxygen consumption for sediment defined as below; SOD usually expressed as a distributed sink term in g m-2 day-1. Sediment oxygen demand is due to

the oxidation of organic matter in bottom sediments. These benthic deposits or sludge beds derive

from several sources. Wastewater particulates, leaf litter and eroded organic rich soil can result in

sediments with high organic content. Regardless of the source, oxidation of the accumulated

organic matter will result in a SOD. Zison et al, (1978) have reported a range of 1.04 to 1.13 for

TSOD. A value of 1.065 is commonly employed [Chapra, 1997].

Equation 27: Oxygen consumption for sediment

( ) ( )

Z

TSOD*SOD

dt

SedDOd 20T−−=

With;

SOD sediment oxygen demand in g m-2 day-1

TSOD temperature coefficient of SOD (1.06)

Oxidation of BOD written as: The analysts introduced sewage sample into a bottle and merely how much oxygen was consumed.

The resulting quantity was dubbed BOD. Water quality analysts early on adopted 5-day BOD test.

Incubation time 5 days makes the test practical, then to extrapolate the 5-day result to the ultimate

BOD level. This is usually done by performing a long term BOD to estimate the decay rate

[Chapra, 1997].

Equation 28: Ultimate BOD

5*Kdexp1

BOD

dt

BODU−−

=

BOD represents the oxygen demand equivalent to the complex biodegradable organic matter

present in water. First order reaction kinetics is used to decay the matter [Mannaerts, 2007].



Equation 29: Biochemical oxygen demand

( ) ( )KNDODO

DO*TKd*BODU*Kd

dt

BODoxd 20T

+−= −

Where;

Kd BOD degradation rate constant in day

BODU ultimate BOD in mg l-1

TKd temperature coefficient of BOD decay (1.05)

KDO Monod constant DO inhibition BOD decay in mg l-1

Nitrification rate solved by the following: Nitrogen compounds in water also have an impact on oxygen resources. Sewage nitrogen can be

broadly broken down into organic nitrogen compounds and ammonia autotrophic bacteria then

assimilate the ammonia and create nitrite and nitrate. The oxidation would consume 4.57 g of

oxygen per 1 g of Kjeldahl nitrogen [Chapra, 1997]. In terms of oxygen required to complete the

nitrification steps form ammonia to nitrate, the equation can be written as:

Equation 30: Oxygen consumption for nitrification

( ) ( )DOKNDO

DO*TKnit*4NH*Knit*57.4

dt

Nitrifd 20T

+−= −

With;

Knit nitrification rate constant in day

NH4 ammonium in mg l-1

TKnit temperature coefficient of nitrification (1.05)

KNDO Monod constant DO inhibition nitrification in mgl-1 [Lijklema, et al., 1996]

5.3: Quality model setup

This water quality model of the Tuul River is a one-dimensional, distributed, mathematical,

conceptual and non-steady state, prototype model. The quality model is linked to the water quantity or

flow model, mentioned in a previous chapter. Therefore, a model network, boundary and initial

conditions are reorganized using output of a flow model. The DO balance of the Tuul River system

has been modelled, calibrated and validated using some existing hydro-chemical data between 2005

and 2006. The following hydro-chemical field datasets were added into a flow model.

1. Monthly data of DO, BOD5, NH4+ at Tuul-Zaisan sampling point

2. Monthly data of DO, BOD5, NH4+ at Tuul-Sonsgolon sampling point

3. Monthly data of DO, BOD5, NH4+ at Tuul-Songino (upper) sampling point

4. Monthly data of DO, BOD5, NH4+ at Tuul-Chicken farm sampling point

5. Monthly data of DO, BOD5, NH4+ at Tuul-Khadanhyasaa sampling point

6. Monthly data of DO, BOD5, NH4+ at Tuul-Altanbulag sampling point

7. Monthly average data of DO, BOD5, NH4+ from the CWTP

8. Hourly average solar radiation data at Ulaanbaatar station downloaded from global open

source

9. Daily average wind speed data from open source

10. Daily average water temperature data at Tuul-Ulaanbaatar station

Rest of data belongs to non-field.



A flowchart has shown general structure and procedure of a quality model.

Figure 39: Flowchart of a quality model

Figure 40: Quality model network schematization



DO modelling network of the Tuul River model consists of the following objects:

− Nodes are representing sampling points along the river

− First 6 sections are demonstrating the Tuul River

− Cross-sections are interpolated along the river

− Two discharge points are indicating CWTP and Bio WTP.

Position of nodes, discharge points and length of the river are all-true. On the other hand, position of

an end node and length of last section are false values that used in a flow model.

To make a quality model out of a flow model, the following steps need to be carried out:

− Define quality description file

− Define initial conditions for quality

− Define boundary conditions for quality

− Define parameters

− Define external variables

− Configure the calculation [DMS, 2004b]

5.3.1: Quality description

A quality description file consists of main two parts:

− Declaration part:

The different variables are defined.

− Compound statement for water courses:

The equations are describing the processes in the water route.

Four types of variables are distinguished in quality description.

1. water quality has effect on this variables.

2. xt external variables, which are space and time dependent.

3. parm parameters, constants and coefficients used in process.

4. flow supplied by the hydraulic part of model [DMS, 2004a].

The following chemical and hydraulic variables were used in DO model:

− WATER DO Dissolved oxygen

− WATER BOD Biochemical oxygen demand

− WATER NH4 Ammonium

− FLOW Q Discharge

− FLOW As Cross-section of flow area

− FLOW Z Water depth

The quality description file, which utilized in this quality model, initially developed in Department

of Water Quality Management and Aquatic Ecology, Agricultural University of Wageningen. See

Appendix 9.

5.3.2: Initial condition of quality model

For all objects, initial concentrations must be defined. Initial values in discharge and level columns

obtained from flow model. Besides, values in chemical columns are collected from field data,

which measured by CLEM.



Table 19: Initial condition in quality model

Model structure Sampling point BOD in mg l-1 DO in mg l-1 NH4+ in mg l-1

Section 1 – begin Tuul-Zaisan 6 9.94 0.7 Section 1 – end Tuul-Sonsgolon 6.1 8.48 0.78 Section 2 – begin Tuul-Sonsgolon 6.1 8.48 0.78 Section 2 – end Tuul-Songino (upper) 6 7.42 1.51 Section 3 – begin Tuul-Songino (upper) 6 7.42 1.51 Discharge point 1 – begin CWTP 30.6 4.41 29.33 Discharge point 1 – end CWTP 30.6 4.41 29.33 Section 3 – end Tuul-Songino (down) 62.8 0.9 1.4 Section 4 – begin Tuul-Songino (down) 62.8 0.9 1.4 Discharge point 2 – begin Bio WTP 62.8 0.9 1.4 Discharge point 2 – end Bio WTP 62.8 0.9 1.4 Section 4 – end Tuul-Chicken farm 26.9 1.14 1.94 Section 5 – begin Tuul-Chicken farm 26.9 1.14 1.94 Section 5 – end Tuul-Khadanhyasaa 9.6 4.4 1.58 Section 6 – begin Tuul-Khadanhyasaa 9.6 4.4 1.58 Section 6 – end Tuul-Altanbulag 6 7.5 1.05 Section 7 – begin Tuul-Altanbulag 6 7.5 1.05 Section 7 – end Extension part 6 7.5 1.05

5.3.3: Boundary condition of quality model

Quality boundary conditions can be entered as time series for concentration. In quality model,

boundary conditions should be defined at nodes and discharge points that have flow boundary

condition.

Table 20: Quality boundary condition

Sampling points

Var

iab

les

in m

g/l

Mar

ch-0

5

Ap

ril-

05

May

-05

Jun

e-0

5

July

-05

Au

gu

st-0

5

Sep

tem

ber

-05

Oct

ob

er-0

5

No

vem

ber

-05

Dec

emb

er-0

5

Jan

uar

y-0

6

Feb

ruar

y-0

6

Mar

ch-0

6

Ap

ril-

06

May

-06

Jun

e-0

6

July

-06

Au

gu

st-0

6

Sep

tem

ber

-06

Oct

ob

er-0

6

No

vem

ber

-06

DO 10 12 6 9 8 9 10 12 n.a n.a n.a n.a n.a 8 8 8 8 9 9 13 13 BOD 6 3 1 3 2 0 1 4 n.a n.a n.a n.a n.a 1 2 1 1 1 1 4 2 Tuul-Zaisan NH4

+ 1 0 0 0 0 0 0 0 n.a n.a n.a n.a n.a 0 0 0 0 0 0 0 0 DO 8 11 8 9 8 11 9 12 11 5 n.a n.a n.a n.a 9 9 9 9 12 11 n.a BOD 6 1 1 3 1 1 1 4 2 2 n.a n.a n.a n.a 2 3 1 1 1 4 n.a

Tuul-Sonsgolon

NH4+ 1 0 0 0 0 0 0 0 0 0 n.a n.a n.a n.a 0 0 0 0 0 0 n.a

DO 7 9 8 9 9 11 11 12 11 n.a n.a n.a n.a 8 9 9 8 9 9 10 11 BOD 6 2 2 2 1 1 2 4 2 n.a n.a n.a n.a 2 2 3 0 1 2 3 1

Tuul-Songino (upper)

NH4+ 2 0 0 0 0 0 1 0 0 n.a n.a n.a n.a 0 0 0 0 0 0 0 0

DO 1 2 9 10 8 9 11 12 6 n.a 4 n.a n.a 3 8 8 6 8 8 8 10 BOD 27 34 2 7 7 5 4 5 35 n.a 21 n.a n.a 52 2 6 2 6 5 14 21

Tuul-Chicken farm

NH4+ 2 12 1 1 1 1 1 2 13 n.a 31 n.a n.a 14 0 1 1 0 1 3 10

DO 4 6 11 9 8 9 11 12 8 3 5 n.a 6 5 8 8 6 8 8 10 8 BOD 10 17 6 8 6 5 3 5 21 9 13 n.a 18 18 4 6 3 5 6 6 4

Tuul-Khadanhyasaa

NH4+ 2 8 1 1 1 1 0 2 6 9 13 n.a 14 9 1 1 0 0 1 0 4

DO 8 10 10 9 8 9 12 12 9 4 n.a n.a 8 4 8 9 6 9 8 11 10 BOD 6 8 4 8 6 3 3 6 3 4 n.a n.a 5 14 6 8 1 5 6 10 2

Tuul-Altanbulag

NH4+ 1 5 1 0 0 1 0 1 5 5 n.a n.a 7 8 1 0 0 0 1 0 4

DO 4 2 4 2 2 2 1 n.a 7 n.a 2 2 1 1 0 2 4 4 3 3 3 BOD 31 24 18 33 31 29 34 32 36 43 49 29 38 52 60 19 20 21 19 27 41 CWTP NH4

+ 29 19 17 16 22 21 21 17 19 16 17 20 20 25 28 35 23 17 30 27 36



Above-mentioned hydro-chemical datasets, which were measured by CLEM and CWTP

respectively, are boundary conditions of nodes and discharge points. Type of boundary condition is

non-equidistant, starts from 01 April 2005 until 30 November 2006. Input data covers only warm

session. All values in boundary conditions are monthly and DMS automatically interpolated

between two values.

5.3.4: Parameters in quality model

In quality declaration section, parameters are abbreviated as “parm”. All constant numbers,

coefficients are belongs to the parameters. The following parameters are used in a prototype

quality model:

− PARM Beta Oxygen production constant (photosynthesis)

− PARM fd Fraction of dissolved BOD

− PARM Kd BOD degradation rate constant

− PARM KDO Monod constant DO inhibition BOD decay

− PARM KLmin Minimum oxygen mass-transfer rate in flowing system

− PARM KNDO Monod constant DO inhibition nitrification

− PARM Knit Nitrification rate constant

− PARM OPTKL Option material transfer

− PARM TKd Temperature coefficient of BOD decay

− PARM TKL Temperature coefficient of oxygen mass-transfer

− PARM TKnit Temperature coefficient of nitrification

− PARM TSOD Temperature coefficient of sediment oxygen demand

− PARM Vs Settling velocity of BOD

5.3.5: External variables

External variables can be defined in the quality model description for input data that are space and,

or time dependent [DMS, 2004b]. It is shortened to “xt”. External variables, which are used in this

model as follow:

− XT T Water temperature

− XT SOD Sediment oxygen demand

− XT SNH4 Diffusion of NH4 concentration

− XT SBOD Diffusion of BOD concentration

− XT I0 Solar radiation

− XT A Algal biomass

− XT W Wind speed

Besides of this, dispersion coefficient has to be described in the quality model itself. The value of

dispersion coefficient can be defined by the following an empirical equation [Chapra, 1997]:

Equation 31: Dispersion coefficient

BS

Q05937.0E

0x =

Where:

Ex dispersion coefficient in m2 s-1

Q average flow in m3 s-1

S0 riverbed slope

B river width in m



Above-mentioned equation proposed by McQuivey and Keefer in 1974. They proposed an

equation the following conditions where the Froude number is less than 0.5, flow ranging from 1

to 934 m3 s-1 and slope from 0.000045 to 0.00298 [Chapra, 1997].

In case of selected section of the Tuul River, average Fr is 0.34, discharge 22.64 m3 s-1 and slope is

0.001981.

Moreover, monthly mean water temperature, was implemented by CLEM and hourly average solar

radiation, monthly mean wind speed, which were measured by Ulaanbaatar weather station,

datasets are obtained from online open course IWEC and applied to this model [ASHRAE, 2001].

5.3.6: Calculation setting

Before running quality model, additional calculation setting should be set. Quality calculation

setting was defined as below:

− Time step is 30 minutes and output is daily

− Type of calculation is “Flow and Quality”

− Theta equals to 0.55

− Output variable is DO

5.4: Calibration

Monthly DO analysis data at sampling point Tuul-Songino (down) was applied to calibrate the quality

model. Chemical data, which used in calibration, measured by CLEM between April and December

2005. The motivation of select this sampling point to calibration, is that the station is the first

sampling point, after the CWTP effluent discharges into the Tuul River. See quality model network.

Figure 41: Correlation between calculated and observed DO in calibration year

R2 = 0.7936

0.00

2.00

4.00

6.00

8.00

10.00

12.00

0.00 2.00 4.00 6.00 8.00 10.00 12.00

Observed Q in m3/s

Cal

cula

ted

Q in

m3

/s

In calibration year, correlation between calculated and observed values was evaluated by different

statistical approaches and the following results calculated in 2005 at Tuul-Songino (down) sampling

point:

− ME -0.76

− MAE 0.90

− RMSE 1.68

− R2 0.79

− E 0.72



Figure 42: Observed and modelled DO in calibration year

0.00

2.00

4.00

6.00

8.00

10.00

12.00

4/1/2005 5/1/2005 6/1/2005 7/1/2005 8/1/2005 9/1/2005 10/1/2005 11/1/2005 12/1/2005

Date

Q in

m3

/s

Observed Modelled

5.4.1: Sensitivity analysis

Sensitivity analyse of the quality model was completed by varying parameters such as β, fd, Kd,

KNDO, Knit, Vs, TKd, TKnit and SOD. In the analysis, minimum and maximum values of

selected parameters were used. Reason of select those parameters, is that the parameters values

have some range in the literature. If model is sensitive to selected parameters then such small

different values give dissimilar statistical results. From those analyses, some selected results are

represented in here. Results of sensitivity analysis were assessed using several statistical methods.

5.4.1.1: Very sensitive parameters

Beta is representing the oxygen production constant by photosynthesis process in the water, due

to algae (and or aquatic plants theoretically). Result of photosynthesis process, certain amount

of oxygen produced by algal biomass. BOD degradation rate constant and nitrification rate

constant are abbreviated as Kd and Knit, respectively.

Figure 43: Sensitivity analysis for beta

0.00

20.00

40.00

60.00

80.00

100.00

120.00

140.00

4/1/2005 5/1/2005 6/1/2005 7/1/2005 8/1/2005 9/1/2005 10/1/2005 11/1/2005 12/1/2005

Date

DO

in m

g/l Observed

B=0.0001

B=0.0005

B=0.001



Table 21: Statistical evaluation of very sensitive parameters

Statistical methods Parameter

values ME MAE RMSE E R2

β=0.0001 -0.76 0.90 1.68 0.72 0.79

β=0.0005 -31.69 31.69 36.01 -130.04 0.25

β=0.001 -70.52 70.52 80.72 -657.48 0.20

Kd=0.1 0.76 0.90 1.68 0.72 0.79

Kd=0.3 -2.43 2.47 2.98 0.11 0.71

Knit=0.1 -2.03 2.03 2.55 0.35 0.79

Knit=0.5 -1.04 1.04 1.90 0.64 0.76

Knit=1.0 -0.76 0.90 1.68 0.72 0.79

Based on above-mentioned graph and table, the model is very sensitive to β, Kd and Knit.

Because of sensitivity, changes have strong effect. The relations between parameters β, Kd and

DO are direct. This means, parameter value increases then DO value increases as well. Between

Knit and DO have inverse relation.

5.4.1.2: Less sensitive parameters

Temperature coefficients of BOD decay and nitrification are shortened as TKd and TKnit,

correspondingly. SOD and fd are short form of sediment oxygen demand and fraction of

dissolved BOD, respectively.

Figure 44: Sensitivity analysis for SOD

0.00

2.00

4.00

6.00

8.00

10.00

12.00

4/1/2005 5/1/2005 6/1/2005 7/1/2005 8/1/2005 9/1/2005 10/1/2005 11/1/2005 12/1/2005

Date

DO

in m

g/l

Observed

SOD=0.5

SOD=2.0



Table 22: Statistical evaluation of less sensitive parameters

Statistical methods Parameter values ME MAE RMSE E R2

TKd=1.03 -0.53 0.90 1.60 0.74 0.78 TKd=1.05 -0.76 0.90 1.68 0.72 0.79

TKnit=1.05 -0.76 0.90 1.68 0.72 0.79 TKnit=1.1 -0.98 1.06 1.94 0.62 0.74 SOD=0.5 -0.98 1.02 1.77 0.68 0.80 SOD=2.0 -0.76 0.90 1.68 0.72 0.79

fd=0.8 -0.84 0.92 1.70 0.71 0.80 fd=0.9 -0.80 0.91 1.69 0.71 0.80 fd=1.0 -0.76 0.90 1.68 0.72 0.79

The model is less sensitive to those parameters that shown in table and graph. Reason of that is

such changes have not strong effect. The relations between parameters TKd, TKnit and DO are

direct. SOD and fd parameters inversely relate to DO value. This means, SOD value increases

then DO concentration reduces.

5.4.1.3: Very less and non sensitive parameters

Monod constant DO inhibition nitrification and settling velocity of BOD are shortened as

KNDO and Vs, respectively.

Figure 45: Sensitivity analysis for Vs

0.00

2.00

4.00

6.00

8.00

10.00

12.00

4/1/2005 5/1/2005 6/1/2005 7/1/2005 8/1/2005 9/1/2005 10/1/2005 11/1/2005 12/1/2005

Date

DO

in m

g/l

Observed

Vs=0.1

Vs=1.0

Table 23: Statistical evaluation of not sensitive parameters

Statistical methods Parameter

values ME MAE RMSE E R2

KNDO=1.0 -0.76 0.90 1.68 0.72 0.79

KNDO=1.5 -0.78 0.91 1.71 0.71 0.79

KNDO=2.0 -0.80 0.92 1.73 0.70 0.78

Vs=0.1 -0.76 0.90 1.68 0.72 0.79

Vs=1.0 -0.76 0.90 1.68 0.72 0.79

Based on results, KNDO is very less sensitive and the model is non-sensitive to Vs. Reason of,

that is Vs was used to calculate BOD concentration.



5.5: Validation

Actual DO measurement, which was implemented by CLEM, applied to validate a prototype quality

model. Validation data covers between April and December 2006.

Figure 46: Correlation between calculated and observed Q data in validation year

R2 = 0.8779

0.00

2.00

4.00

6.00

8.00

10.00

0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00

Observed Q in m3/s

Cal

cula

ted

Q in

m3/s

In validation year, correlation between calculated and modelled values was evaluated by different

statistical approaches and the following results calculated in 2006 at Tuul-Songino (down) sampling

point:

− ME -0.01

− MAE 0.45

− RMSE 0.87

− R2 0.88

− E 0.87

Figure 47: Observed and modelled Q in validation year

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.00

10.00

4/1/2006 5/1/2006 6/1/2006 7/1/2006 8/1/2006 9/1/2006 10/1/2006 11/1/2006 12/1/2006

Date

Q in

m3

/s

Modelled Observed



5.5.1: Scenarios for SWQ improvement

The quality model has been used to evaluate the impact of scenarios for improvement of water

quality in order to meet the Mongolian National Standard. The national standard declares that,

permissible level of DO concentration must equal or above 6 mg l-1 in warm session [CSM, 1998].

A threshold value of 6 mg l-1 was selected in this scenario. Tuul-Songino (down) sampling point

(same as calibration and validation) was selected for checking point and is close to the discharge

point from the CWTP.

If assuming, river discharge and DO, BOD5, NH4+ concentration in the upstream reach, before

CWTP effluent pours into the Tuul River, to be the same as in 2005 at Tuul-Songino (upper)

sampling point, the following options can be suggested:

− Option 1 for ammonium: DO =< 4 mg l-1, BOD5 => 17 mg l-1, NH4+ => 8 mg l-1

− Option 2 for BOD: DO =< 4 mg l-1, BOD5 => 17 mg l-1, NH4+ => 15 mg l-1

− Option 3 for DO: DO =< 8 mg l-1, BOD5 => 17 mg l-1, NH4+ => 15 mg l-1

The concentration values in the options were taken from minimum and maximum concentration

values of CWTP discharge in 2005. That means, CWTP discharge contained NH4+

min=15.74 mg l-1,

DOmax=6.93 mg l-1, BODmin=17.89 mg l-1 in 2005. In other words, the CWTP has a “serious”

possibility to achieve this point.

Figure 48: Options for SWQ improvement

0

2

4

6

8

10

12

14

4/1/

05

5/1/

05

6/1/

05

7/1/

05

8/1/

05

9/1/

05

10/

1/0

5

11/

1/0

5

12/

1/0

5

Date

DO

in m

g/l

O1; O=4, B=17, N=8

O2; O=4, B=17, N=15

O3; O=8, B=17, N=15

Threshold value

Monthly average O1

Monthly average O2

Monthly average O3

Observed in 2005

Option 1:

If CWTP discharge contains DO =< 4 mg l-1, BOD => 17 mg l-1, and NH4+ => 8 mg l-1 then DO in

the Tuul River improves as below:

Table 24: Scenario for SWQ improvement, Option 1

Month April May June July August September October November Simulated 8.7 7.7 9.3 10.9 12.8 12.3 10.8 7.9 Observed 2.6 8.4 9.3 8.5 9.8 10.7 10.2 4.0 Difference +6.1 -0.7 0 +2.4 +3.0 +1.6 +0.6 +3.9

Relative difference, % -234.6 +8.3 0 -28.2 -30.6 -15.0 -5.9 -97.5 RMS 3.0



Option 2:

If CWTP discharge contains DO =< 4 mg l-1, BOD => 17 mg l-1, and NH4+ => 15 mg l-1 then DO

in the Tuul River improves as below:


Month April May June July August September October November Simulated 8.5 6.9 8.8 10.2 11.9 11.5 9.9 6.6 Observed 2.6 8.4 9.3 8.5 9.8 10.7 10.2 4.0 Difference +5.9 -1.5 -0.5 +1.7 +2.1 +0.8 -0.3 +2.6

Relative difference, % -226.9 +17.9 +5.4 -20.0 -21.4 -7.5 +2.9 -65.0 RMS 2.6

Option 3:

If CWTP discharge contains DO =< 8 mg l-1, BOD => 17 mg l-1, and NH4+ => 15 mg l-1 then DO

in the Tuul River improves as below:


Month April May June July August September October November Simulated 9.5 7.1 8.9 10.3 12.0 11.6 10.1 6.8 Observed 2.6 8.4 9.3 8.5 9.8 10.7 10.2 4.0 Difference +6.9 -1.3 -0.4 +1.8 +2.2 +0.9 -0.1 +2.8

Relative difference, % -265.4 +15.5 +4.3 -21.2 -22.4 -8.4 +1.0 -70 RMS 2.9

Based on these figure, tables, suggested values and statistical analysis, the best option is the first.

Because of, monthly mean DO concentrations are not lower than the threshold value and monthly

improvement are greater than other options. Those options are valid only for the warm session.

Obviously, discharge from CWTP should contain DO concentration greater and BOD5, NH4+

concentrations less than the selected option in cold session.

5.6: Chapter conclusion

Some conclusions drawn from the quality modelling are mentioned as below:

− Quality model syntax (oxygen model) can be used for this river.

− Very limited number of observations (once per month) has an effect on the results, such as

high statistical results.

− In sensitivity analysis, the model is very sensitive to β, Kd and Knit. Parameters TKd, TKnit,

SOD and fd are less sensitive. Parameter KNDO is very less and Vs is not affective to the

model. Relations between DO concentration and β, Kd, KNDO, TKd, TKnit are direct.

Inverse relations have between DO and fd, Knit, SOD.

− Best option for SWQ improvement is first (see before). Because, monthly mean DO

concentrations are not lower than the threshold value and monthly improvement are greater

than other options.

5.7: Limitation in quality modelling

A relatively limited number of chemical analysis data makes it difficult to simulate and to make more

detailed water quality modelling and analysis. The source data quality is not always clear due to

miscommunication and weak cooperation between water issues implementing organizations.



Chapter 6: Conclusions and recommendations

6.1: Overall conclusions

A brief sequence of entire study as implemented is described as below:

− Proposal writing

− Organization and execution of field campaign

− Surface water quality assessment

− River flow model building and verification

− Water quality model building and analysis

− Conclusion and recommendations

A thesis proposal was written, based on real-time problems in Mongolia and mainly focussing on

water quality assessment, pollution mapping and modelling.

A field campaign was executed in the Ulaanbaatar city, Mongolia. Most vital information, data and

water samples were collected and analysed during the field work.

A method using a surface water quality index was utilized in assessment part. Six parameters i.e.

dissolved oxygen, ammonium, nitrate, nitrite, biochemical oxygen demand and chemical oxygen

demand are involved in index calculation. ILWIS 3.4 was used to visualise time series of thematic

water quality maps.

Before the surface water quality model development, a river flow model was developed using DMS

model code and using field observation. Totally, 24 chemical state variables and parameters took part

in quality model for dissolved oxygen.

More detailed specific conclusions are drawn at the end of chapters 3, 4 and 5. Overall conclusions,

recommendations, future research, output significance of study and limitations, difficulties, which

were encountered during study, are detailed in this chapter.

From this study, the following conclusions can be drawn:

− The general trend of water quality of the Tuul River showed a significant decrease throughout

the 11-year analysis period (1996-2006). Especially the downstream section of the Tuul

River, downstream of the Ulaanbaatar city, showed a decreasing trend in water quality. In the

year 1999, the water quality was comparatively good, with as main reason, new equipments

installation in the CWTP using JICA financial support. In the 2005, the Tuul River got

strongly polluted caused by poor operation of the CWTP.

− The water pollution of the Tuul River is gradually reducing along the downstream reach, but

not complete self-purification is reached at the Tuul-Altanbulag sampling point (35 km

down).

− Cold period negatively affects to the self-purification capacity and processes of the river.

Another reason is that discharges from pollutant sources are in this period greater than river

discharges.

− The CWTP remains the largest, strongest point pollution source in this branch of the Tuul

River. Political, scientific, engineering groups and water issues implementing organizations



have to pay more attention to the functioning of this treatment plant. Recent evidence

suggests that a main problem of this CWTP is reliable electricity supply.

− DMS software and existing quality model syntax can be used in this selected section of the

Tuul River.

− If real water level value or riverbed values are greater than one thousand, then DMS shown an

error message. Maybe, this is a bug of this software. This can be compensated by

substraction.

− Very limited number of chemical observations (once per month) can lead to significant

observation and model uncertainties. For example, sampled and analysed day and time is not

clear, only one analysis cannot always represent an entire month.

− When comparing the two models, a quality model (oxygen balance) is more sensitive than a

water flow and quantity model. Specially, the parameters β, Kd, Knit are very sensitive. Beta,

Kd, TKd, TKnit, KNDO have direct and Knit, SOD, fd have inverse relation to dissolved

oxygen concentration, respectively.

− Based on three developed scenarios, best one for SWQ improvement is first option. Because

of, monthly mean DO concentrations are not lower than the threshold value and monthly

improvement are greater than other options.

6.2: Recommendations

The pollution in the Tuul River is a serious problem, which needs to be addressed, urgently. A

different approach is suggested and needs to be taken in order to solve the complex issues involved.

� In Mongolia, many organizations are involved in water quantity and quality issues. Result of

that is miscommunication and data quality is unfavourable. In 2003, the government decided

to create a new organization, the Water Authority, which has the responsibility for water

resources issues. However, this organization still could not be involved in all urgent water

issues, such as the Tuul River water quality. Without any delay, the Water Authority should

take responsible for all urgent water issues in one centralized body and reorganize all other

small organizations under law.

� Another important issue is that a river basin management system needs to be developed for

the Tuul River basin. The NGIC project, which is funded by Dutch government, is trying to

develop such a management system. This project started in 2006 and will continue until 2009.

� The CWTP has to regularly renew its equipments (e.g. pumps, filtration, etc.), urgently and improve efficiency of the operation system, keep “good” controls in the operational system, constantly. In that case, maybe the CWTP needs to build its own independent electric power station.

� To implement a project for quality modelling, organizations in water issues should be involved. In the model, we should use datasets, which covers longer and real-time, sampled at higher time frequency and maybe also more sampling points. The wastewater carrying capacity of the Tuul River could then be estimated based on the model.



6.3: Future research

The result of this study is not the end of this research. In future, we need to update some of the data

and improve the obtained results. For example:

− To change type of cross-section from trapezoid to asymmetric, and to improve the detailed

physical schematizations of the river network and pollution outfalls;

− To extend the entire study period in time;

− To collect and use more detailed data, that represents shorter time intervals and periods. This

means, chemical datasets, used in quality model, should be monitored (e.g. automatically) on

a daily basis rather than at a monthly scale. For more accurate result, we need to use shorter

time interval data;

− The spatial network could also be refined and more detailed, and the distances between nodes

could be decreased.

− Due to the elevation of the Ulaanbaatar city and the Tuul River, we should include air

pressure effects and water salinity data in oxygen saturation calculation.

− Although the Tuul River flows main origins are rainfall based and snowmelt streamflow, the

DMS model for both quantity and surface water quality could be set to interact with the

shallow groundwater using Moduflow option in DMS. Also could the DMS model be

extended with a rainfall-runoff of RAM component, to represent the whole water balance of

the river basin. This will require a future research and application project.

6.4: Output significance

Until now, this research has the following importance.

− The maps and assessment of river water contamination and a water quality model are useful

for decision makers and daily human life, well-being.

− In addition, outputs of research can be informative for urban planning.

− This research can be the basis of future research of river water contamination using different

spatial and temporal datasets.

6.5: Limitations of the study

In this study, the following limitations were faced:

− Some hydraulic and chemical datasets were very limited, generalized and not accessible by

internet. For example, cross-section data were not available. According to some information,

real cross-sectional data are not measured at Tuul-Altanbulag station.

− Most of data are costly and some of data still belonged to the national secret.

− The time to produce this MSc research was short and limited (only 6 months)

− Very limited number of chemical analysis data made it became difficult to simulate and to

make more detailed quality analysis. Data quality was not always clear due to

miscommunication and general weak cooperation between water issues implementing

organizations.



− Correlation between water quality and quantity could not be studied in detail. Reason of that

is the Tuul River hydraulic parameters measurement and quality analysis are implemented by

two different organizations. Therefore, measurement dates and locations of both organizations

are not coinciding, which makes combined use very difficult. For example, daily discharge

measures in three different stations, that are belong to WS, NAMHEM, are situated in study

area. Ones in each month, the water quality analysis is implemented by CLEM in different

locations from the hydraulic measurements. Furthermore, the water quality analysis day of

sampling or observation is not written into the chemical dataset provided by CLEM.

− In the flow model simulation, use of actual water level data in both stations, especially at

Tuul-Altanbulag station, was sometimes confusing due to the implementation of water level

measurements using artificial reference surface level, instead of a permanent ground

reference.



References

Adiyabadam, G. (1996), Chemical content of river, contamination rank and its model, National University of Mongolia, Ulaanbaatar.

Altansukh, O. (2000), Surface water pollution of Ulaanbaatar city, 43 pp, National University of Mongolia, Ulaanbaatar.

ASHRAE (2001), International Weather for Energy Calculations (IWEC weather files), edited, ASHRAE.

Baasandorj, D., and G. Davaa (2006), Tuul river basin of Mongolia - Integrated water resource management, First edition ed., Interpress, Co.,LTD, Ulaanbaatar.

Batima, P. (1998), Chemical content of Mongolian rivers and its quality assessment, National University of Mongolia, Ulaanbaatar.

Chapra, S. C. (1997), Surface water quality modelling, 843 pp., McGraw-Hill, Boston etc.

CSM (1998), Standard for surface water quality, edited, p. 30, Centre of Standardization and Measurement.

CWTP (2006), Annual report of CWTP, presentation, Ulaanbaatar.

Dashdeleg, N., and B. Bat (1971), Mongolian rivers, Ulaanbaatar.

Davaa, G., et al. (2006), Surface water of Mongolia, Environmental book of Mongolia, 55-82.

Dingman, S. L. (2002), Physical hydrology, Second edition ed., 646 pp., Prentice Hall, Upper Saddle River.

DMS (2004a), Duflow manual, edited, Duflow modelling studio.

DMS (2004b), User's guide, edited, Duflow Modelling System.

DMS (2007), Duflow Modelling System, edited, p. The official website of DMS, Delft, The Netherlands.

Dost, R. J. J. (2006), Water quality sampling and testing field guide, 1 ed., 136 pp., ITC, Enschede.

Edwards, K. (1998), Manning's Equation Calculator / Software, edited, LMNO Engineering, Research, and Software, Ltd.

Furniss, M., et al. (2006), Froude number and flow states, edited, USA.

Helsel, D. R., and R. M. Hirsch (1992), Statistical methods in water resources, 522 pp., Elsevier, Amsterdam etc.

Javzan, C., et al. (2004), Study of Tuul river contamination, Geo-ecology in Mongolia, 04, 213-219.



Krenkel, P. A., and V. Novothy (1980), Water quality management, 671 pp., Academic Press, New York.

Lijklema, L., et al. (1996), Model for the DO in water system, in DMS, edited, Department of Water Quality Management and Aquatic Ecology, University of Wageningen, Wageningen.

Makkinga, A. A., et al. (1998), Modelling dissolved oxygen in the river Regge system using Duflow, European Water Management, 1(1998)1, pp. 37-42, 1, 6.

Mannaerts, C. (2007), The oxygen balance in rivers, in Water quality modelling, edited, Enschede, The Netherlands.

Merwade, V. (2005), Distributed flow routing, edited.

MNE (1997a), Causes of decreasing water resource of Tuul River and ways of its protection, Ministry of Nature and Environment, Ulaanbaatar.

MNE (1997b), Definition of quality grades used to surface water in Mongolia, edited, p. 14, Ministry of Nature and Environment.

MNE (2006), Report on State of Environment for 2004-2005, edited, p. 27, Ministry of Nature and Environment, Ulaanbaatar.

Munguntsetseg, A. (1987), Chemical content and contamination of Tuul river, National University of Mongolia, Ulaanbaatar.

Munguntsetseg, A., et al. (1982), Chemical component and pollution of Tuul River, National University of Mongolia, Ulaanbaatar.

NAMHEM (1999), Surface water in Mongolia, First edition ed., Interpress Co.,LTD, Ulaanbaatar.

Nash, J. E., and J. V. Sutcliffe (1970), River flow forecasting through conceptual models part I -- A discussion of principles, Journal of Hydrology, 10, 282-290.

Orchlon (1995), Pollution mitigating measures for the Tuul river - Three assessments, in Daily news, edited, Ulaanbaatar.

Orgil, B. (2007), Conceptual approach for water sub component of NGIC project, edited by M. o. N. a. Environment, p. 15, NGIC for NRM project.

Oyunchuluun, Y. (1999), Assessment of water quality using river bed insects, Pedagogical University of Mongolia, Ulaanbaatar.

Patankar, S. V. (1980), Numerical heat transfer and fluid flow, Taylor & Francis, [S.l.].

Rientjes, D. T. H. M. (2006), Modelling in hydrology, 250 pp., ITC, Department of Water Resources, Enschede, The Netherlands.

Rossiter, D. G. (2007), Using the R Environment for Statistical Computing, in Applied geo-statistics, edited, p. 44, International Institute for Geo-information Science & Earth Observation (ITC), Enschede.

Roza-Butler, N. (2004), An overview of the current condition of the Tuul river, Geo-ecology in Mongolia, 04, 220-226.



Soninhishig, N. (1998), Algae in Tuul river, National University of Mongolia, Ulaanbaatar.

SRTM (2005), Shuttle Radar Topography Mission, edited, USA.

Thomann, R. V., and J. A. Mueller (1987), Principles of surface water quality modelling and control, Harper Collins, New York.

Tuvaanjav, G. (1972), Water regime of Tuul river, Journal for Mongolian geography, 11.

Tuvaanjav, G. (1978), Tuul River, hydrochemistry of its tributaries, Journal for Mongolian geography, 18.

Tuvaanjav, G. (1983), Changes of natural water quality and component near by cities, Journal for Mongolian geography, 23.

Uranbileg, L. (2000), River pollution in surrounding area of Ulaanbaatar city, in Conference for ecological problems in Mongolia, edited, Ulaanbaatar, National University of Mongolia.

White, F. M. (1998), Fluid Mechanics 4th Bk&Dk edition ed., 294 pp., Mcgraw-Hill College.

WHO (1996), Water quality assessments: A guide to the use of biota, sediments and water in environmental monitoring, Second edition ed., 626 pp., Spon, London etc.

Zagas, B. (1998), Assessment of water quality using water insects, National University of Mongolia, Ulaanbaatar.



Appendices

Appendix 1: Photos in the field





Appendix 2: Note for water field survey

A Information of sampling location

1 Name of state, province and municipality Mongolia, Ulaanbaatar,

2 Name of water body Tuul

3 Name of basin and sub-basin Tuul river basin and ……………………......... sub-basin

4 Sampled medium River

5 Sampling location

6 Geographical coordinate N(y)

E(x)

7 Elevation of sampling location ……………………………. meter a.s.l

8 Sampling site description

9 Water turbulence grade weak moderate high

10 Land use around sampling location

11 Possible contamination source

B Sample information 12 Name of collector Ochir Altansukh

13 Collected date and time 2007 ………………………………

14 Sample number or name

15 Sample container ……………… ml ……………..

16 Sampling method

standard method MNS ISO 5667-6 : 2001 “Water quality.

Sampling, 6th part. Guideline for sampling from river and

stream”

17 Depth of sampling middle of water depth

18 Water temperature …………….. 0C

19 Preservation method, if any

20 Identification of project MSc research project

21 Purpose of sampling water quality assessment

22 Weather condition during sampling

23 Any changes from normal weather

condition

24 Any specific observation in-situ

C Measurement result 25 Variable measured

26 Actual results

27 Measured unit

28 Analytical method

29 Measured place in-situ field laboratory standard laboratory

30 Name of instrument and laboratory



Appendix 3: Field measurements

NOTE OF FIELD SURVEY Measured by Mr. Ochir Altansukh (2007.09.17-22)

Geographical coordinate Measured variable

N River name Ex Ny

Elevation Water

turbulence 0C pH DO in mg l-1 EC in µµµµS cm-1

HACH file name

GPS point name

Photo file

1 Terelj - Terelj 47059/30.0// 107027/36.1// 1522 weak 7.9 7.59 8.73 49.6 Terelj 001 - 002 Ter - 01 01 - Terelj 01

2 Terelj - UB-2 tourist camp 47059/16.8// 107027/53.8// 1521 weak 9.3 7.47 7.78 52.4 Terelj 003 - 004 Ter - 02 02 - Terelj 02

3 Terelj - 03 47058/21.5// 107028/43.1// 1507 weak 10.1 7.37 7.71 51.5 Terelj 005 - 006 Ter - 03 none

4 Tuul - Uubulan 47048/26.4// 107022/50.3// 1383 weak 13.4 7.96 8.02 56.9 Terelj 007 - Tuul 001 T - 001 03 - Tuul 01

5 Tuul - Terelj bridge 47049/20.6// 107020/02.3// 1375 moderate 12.8 8.03 8.10 58.8 Tuul 002 - 003 T - 002 04 - Tuul 02

6 Tuul - 03 47049/43.5// 107019/25.0// 1372 moderate 12.7 8.04 8.07 59.3 Tuul 004 - 006 T - 003 05 - Tuul 03

7 Tuul - Nalaih 47049/14.0// 107015/56.4// 1364 weak 11.6 8.23 8.81 64.7 Tuul 007 - 008 T - 004 06 - Tuul 04

8 Nalaih - 01 47048/36.7// 107016/43.9// 1382 weak 16.1 7.36 6.15 768.0 Tuul 009 - 010 N - 001 07 - Nalaih 01

9 Nalaih - 02 47047/52.0// 107016/26.9// 1397 weak 18.9 8.57 7.43 778.0 Tuul 011 - 012 N - 002 08 - Nalaih 02

10 Nalaih - WWTS 47046/46.5// 107015/50.2// 1426 sewage 15.2 7.69 6.68 523.0 Tuul 013 - 014 N - 003 09 - WWTS 01

11 Nalaih - Spring 47046/46.5// 107015/50.2// 1426 spring water 14.3 8.51 8.85 1206.0 Tuul 017 - 018 N - 003 10 - Nalaih 03

12 Confluence of Nalaih spring and sewage 47046/50.6// 107015/51.8// 1422 weak 16.3 7.95 5.27 580.0 Tuul 019 - 020 N - 004 none

13 Tuul -05 47053/06.7// 107014/50.5// 1350 moderate 13.2 8.54 8.45 280.0 Tuul 021 - 022 T - 005 11 - Tuul 05

14 Tuul -06 47053/45.2// 107012/50.3// 1352 moderate 12.9 8.21 8.31 398.0 Tuul 023 - 024 T - 006 12 - Tuul 06

15 Tuul - Gachuurt 47054/28.1// 107010/40.9// 1340 weak 12.0 8.15 8.24 132.0 Tuul 025 - 026 T - 007 13 - Tuul 07

16 Tuul - Khar usan tohoi 47054/39.9// 107007/44.0// 1334 moderate 9.3 7.95 8.88 244.0 Tuul 027 - 028 T - 008 14 - Tuul 08

17 Tuul - Bayanzurh bridge (upstream) 47053/55.1// 107004/12.8// 1313 moderate 9.4 8.06 8.81 165.1 Tuul 029 - 030 T - 009 15 - Tuul 09

18 Tuul - Bayanzurh bridge (downstream) 47053/28.1// 107003/04.7// 1309 weak 11.5 8.78 9.21 147.8 Tuul 031 - 032 T - 010 16 - Tuul 10

19 Tuul - Railway bridge (downstream) 47053/36.7// 107001/55.0// 1305 weak 14.0 7.98 7.91 335.0 Tuul 033 - 035 T - 011 17 - Tuul 11

20 Tuul - 12 47053/30.5// 106059/45.2// 1297 weak 15.8 8.31 8.57 223.0 Tuul 036 - 037 T - 012 18 - Tuul 12

21 Tuul - Zaisan 47053/19.4// 106055/05.7// 1306 weak 15.0 8.57 8.23 115.8 Tuul 038 - 039 T - 013 19 - Tuul 13

22 Tuul - 14 47053/26.6// 106052/48.6// 1276 moderate 15.4 8.48 8.40 134.2 Tuul 040 - 041 T - 014 20 - Tuul 14

23 Tuul - Niseh bridge (downstream) 47053/17.4// 106050/47.3// 1276 weak 14.9 8.37 8.18 231.0 Tuul 042 - 043 T - 015 21 - Tuul 15

24 Selbe - Khandgait 48007/03.7// 106053/27.7// 1497 weak 9.5 8.09 8.70 224.0 Tuul 044 - 045 S - 001 22 - Selbe 01

25 Khandgait - 01 48006/44.7// 106054/07.4// 1505 weak 12.1 8.10 8.74 309.0 Selbe 001 - 002 S - 002 23 - Khandgait 01

26 Selbe - Belh bridge 48004/40.3// 106053/49.1// 1443 weak 13.2 8.37 8.25 267.0 Selbe 003 - 004 S - 003 24 - Selbe 02

27 Selbe - Sharga Morit 48002/58.7// 106054/10.7// 1411 moderate 14.4 8.67 8.40 249.0 Selbe 005 - 006 S - 004 25 - Selbe 03



28 Selbe - Shadivlan 48001/00.2// 106055/16.9// 1381 weak 15.9 8.61 8.01 310.0 Selbe 007 - 008 S - 005 26 - Selbe 04

29 Selbe - Dambadarjaa bridge 47058/36.8// 106055/32.9// 1346 weak 17.9 9.50 8.45 282.0 Selbe 009 - 010 S - 006 27 - Selbe 05

30 Selbe - Selbe bridge 47055/44.3// 106055/53.3// 1303 weak 18.1 9.06 9.98 424.0 Selbe 011 - 012 S - 007 28 - Selbe 06

31 Selbe - Lion bridge 47055/05.0// 106055/45.9// 1304 weak 18.5 9.13 9.41 508.0 Selbe 013 - 014 S - 008 29 - Selbe 07

32 Mouth of Tolgoit stream 47053/51.6// 106049/46.8// 1280 weak 17.9 8.24 10.06 678.0 Selbe 015 - 016 To - 001 30 - Tolgoit 01

33 Tuul - Sonsgolon 47052/44.6// 106048/01.8// 1256 moderate 16.3 8.80 7.99 110.6 Tuul 046 - 047 T - 016 31 - Tuul 16

34 Confluence of Tuul and Selbe 47053/01.7// 106048/47.5// 1266 weak 16.3 8.77 7.76 320.0 Tuul 048 - 049 T - 017 32 - Tuul 17

35 Tuul - 18 47053/01.7// 106048/47.5// 1266 weak 16.2 8.76 7.80 75.6 Tuul 050 - 051 T - 017 33 - Tuul 18

36 Mouth of Selbe stream 47053/01.7// 106048/47.5// 1266 weak 16.0 8.76 8.27 527.0 Selbe 017 - 018 T - 017 34 - Selbe 08

37 Selbe - 09 47053/27.9// 106049/19.5// 1261 weak 15.9 8.30 3.54 540.0 Selbe 019 - 020 S - 009 35 - Selbe 09

38 Confluence of Selbe and Tolgoit 47053/47.1// 106049/42.7// 1270 weak 15.9 8.32 9.69 690.0 Selbe 021 - 022 S - 010 36 - Selbe 10

39 Selbe - 11 47053/47.1// 106049/42.7// 1270 weak 13.6 7.38 2.82 418.0 Selbe 023 - 024 S - 010 37 - Selbe 11

40 Tuul - Sonsgolon bridge (downstream) 47052/28.7// 106046/50.1// 1262 moderate 12.0 8.26 8.70 77.6 Tuul 052 - 053 T - 018 38 - Tuul 19

41 Tuul - Niseh 47052/01.6// 106043/05.9// 1246 moderate 13.1 8.41 8.74 468.0 Tuul 054 - 055 T - 019 39 - Tuul 20

42 WWTP - Railway bridge 47052/13.5// 106041/35.6// 1239 weak 17.8 7.73 0.62 1134.0 Tuul 056 - 057 T - 020 40 - WWTS 02

43 Tuul - Songino resort (upper) 47052/01.8// 106041/42.8// 1238 moderate 13.6 8.39 8.64 81.0 Tuul 058 - 059 T - 021 41 - Tuul 21

44 Confluence of Tuul and WWTP 47052/01.8// 106041/42.8// 1238 weak 14.5 7.96 7.21 213.0 Tuul 060 - 061 T - 021 none

45 Confluence of Tuul and WWTP (30 m in downstream) 47051/52.0// 106041/32.2// 1226 weak 15.6 7.82 6.42 281.0 Tuul 062 - 063 T - 022 42 - Tuul 22

46 Tuul - Bio bridge 47050/51.7// 106040/29.7// 1235 moderate 13.9 7.95 7.19 242.0 Tuul 064 - 065 T - 023 43 - Tuul 23

47 Tuul - Songino (upper) 47051/17.8// 106041/23.2// 1237 weak 14.3 8.50 9.05 135.2 Tuul 066 - 067 T - 024 44 - Tuul 24

48 Discharge from WWTS of Bio 47050/39.0// 106040/12.7// 1248 discharge 16.1 7.48 6.14 177.2 Tuul 068 - 069 T - 025 45 - WWTS 03

49 Tuul - Chicken farm (upstream) 47046/21.0// 106035/59.2// 1211 weak 15.7 8.41 8.13 238.0 Tuul 070 - 071 T - 026 46 - Tuul 25

50 Tuul - Khadanhyasaa (Salhitiin Gatsaa) 47045/08.9// 106030/02.6// 1205 weak 9.4 7.72 6.75 422.0 Tuul 072 - 073 T - 027 47 - Tuul 26

51 Tuul - Altanbulag 47042/59.8// 106024/29.5// 1180 weak 10.6 7.58 7.41 305.0 Tuul 074 - 075 T - 028 48 - Tuul 27

52 Tuul - Altanbulag (Guurnii Gatsaa) 47041/53.4// 106017/40.6// 1185 weak 16.0 8.76 10.05 274.0 Tuul 076 - 077 T - 029 49 - Tuul 28



Appendix 4: Hydro-chemical dataset

Chemical monthly data of the Tuul river and its tributaries

EC DO NH4+ NO2

- NO3-

Total mineral N

COD-Mn BBOODD Date pH

µS cm -1 mg l -1 mg l -1 mg l -1 mg l -1 mg l -1 mg l -1 mg l -1

Tuul - Songino (down) January-96 7.5 571.0 7.37 15.30 0.066 1.42 16.786 6.5 14.0

February-96 n.a n.a n.a n.a n.a n.a n.a n.a n.a March-96 8.6 389.0 5.93 15.80 0.078 1.77 17.648 11.1 14.3

April-96 7.8 590.0 3.95 1.04 0.149 0.88 2.069 3.6 May-96 7.8 174.4 7.90 1.18 0.042 0.69 1.912 3.7 4.0

June-96 8.0 268.0 9.42 2.12 0.063 1.28 3.463 3.5 0.2 July-96 7.7 135.3 8.78 0.29 0.022 0.50 0.812 4.0 6.8

August-96 7.4 135.9 9.24 0.51 0.034 0.32 0.864 4.1 2.9 September-96 7.6 164.7 8.26 1.54 0.040 0.47 2.050 2.4 1.6

October-96 7.6 237.0 8.06 0.83 0.037 0.66 1.527 2.9 4.3 November-96 7.6 504.0 3.12 11.40 0.151 2.13 13.681 4.4 10.7 December-96 n.a n.a n.a n.a n.a n.a n.a n.a n.a

Annual average -96 7.8 316.9 7.20 5.00 0.068 1.01 6.081 4.6 6.5 January-97 7.9 565.0 2.61 n.a n.a n.a n.a 3.7 12.46<

February-97 8.0 545.0 2.13 18.90 0.069 0.96 19.929 5.3 37.7 March-97 7.9 437.0 4.10 0.925 0.925 n.a 3.5

April-97 7.7 207.0 8.16 1.60 0.047 0.69 2.337 4.1 7.3 May-97 7.7 270.0 8.16 3.94 0.108 0.81 4.858 2.9 8.16<

June-97 n.a n.a n.a n.a n.a n.a n.a n.a n.a July-97 7.9 96.6 10.75 0.24 0.032 0.41 0.682 4.8 3.1

August-97 7.7 81.9 8.06 0.50 0.014 0.514 4.0 1.7 September-97 7.9 144.7 9.36 1.11 0.024 0.29 1.424 2.6 2.9

October-97 8.0 227.0 9.07 2.39 0.079 0.92 3.389 3.5 6.9 November-97 8.3 399.0 4.68 8.92 0.187 1.02 10.127 4.1 15.56< December-97 8.5 456.0 4.97 12.70 0.135 1.71 14.545 5.3 18.7

Annual average -97 7.9 311.7 6.55 5.59 0.162 0.85 5.873 4.0 10.2 January-98 n.a n.a n.a n.a n.a n.a n.a n.a n.a

February-98 n.a n.a n.a n.a n.a n.a n.a n.a n.a March-98 8.7 502.0 n.a 9.59 0.313 1.43 11.333 3.6 20.8

April-98 7.2 n.a 9.09 2.24 0.832 2.28 5.352 5.4 13.4 May-98 7.0 217.0 9.16 2.28 0.250 0.63 3.160 3.2 4.8

June-98 n.a n.a n.a n.a n.a n.a n.a n.a n.a July-98 7.7 109.5 7.37 1.05 0.035 0.23 1.315 3.0 2.7

August-98 n.a n.a n.a n.a n.a n.a n.a n.a n.a September-98 7.3 85.2 8.74 0.28 0.007 0.57 0.857 5.0 1.9

October-98 7.3 168.9 11.32 0.67 0.175 0.56 1.405 3.1 6.6 November-98 6.7 330.0 9.42 3.26 0.003 1.09 4.353 4.9 7.0 December-98 7.4 360.0 6.69 10.20 0.124 1.92 12.244 5.9 8.06<


February-99 n.a n.a n.a n.a n.a n.a n.a n.a n.a March-99 7.8 470.0 9.27 11.74 0.028 0.89 12.658 5.5 5.4

April-99 7.3 63.8 10.84 0.46 0.007 0.20 0.667 6.6 2.0 May-99 7.9 110.4 11.08 0.81 0.020 0.31 1.140 3.9 4.8

June-99 6.9 63.4 8.27 0.32 0.010 0.30 0.630 7.3 3.0 July-99 8.1 80.7 7.80 0.30 0.110 0.18 0.590 4.6 1.4




February-00 n.a n.a n.a n.a n.a n.a n.a n.a n.a March-00 n.a n.a n.a n.a n.a n.a n.a n.a n.a

April-00 8.0 250.0 8.80 4.25 0.169 1.41 5.829 4.5 7.9 May-00 7.4 244.0 6.60 3.67 0.070 0.41 4.150 13.0 6.6

June-00 8.2 195.4 8.70 2.52 0.073 0.80 3.393 3.2 5.1 July-00 8.3 141.0 7.90 1.05 0.034 0.35 1.434 2.5 2.3


October-00 7.6 349.0 9.50 6.26 0.216 0.72 7.196 6.6 8.3 November-00 n.a n.a n.a n.a n.a n.a n.a n.a n.a December-00 7.6 617.0 6.98 11.82 0.219 1.01 13.049 5.5 12.7

Annual average -00 7.9 254.4 8.39 3.93 0.105 0.68 4.707 4.9 6.1 January-01 8.0 563.0 4.47 13.06 0.282 1.06 14.402 8.3 10.8

February-01 7.6 543.0 8.62 14.00 0.322 3.65 17.972 5.7 13.6 March-01 n.a 685.0 7.01 18.52 0.167 1.52 20.207 6.6 6.1

April-01 7.9 195.0 9.23 1.66 0.011 0.47 2.141 12.3 6.0 May-01 6.9 122.0 8.18 1.42 0.008 0.33 1.758 8.2 3.3

June-01 n.a 199.0 5.68 0.09 0.003 3.00 3.093 4.6 5.7



July-01 7.2 207.0 7.18 3.13 0.679 0.36 4.169 4.9 6.3 August-01 7.3 115.0 7.85 1.12 0.048 0.19 1.358 5.1 4.6

September-01 7.6 173.0 7.90 1.78 0.025 1.30 3.105 4.0 4.3 October-01 8.0 300.0 7.30 4.53 0.049 0.57 5.149 6.1 7.3

November-01 8.0 681.0 2.74 2.55 0.010 0.11 2.670 18.5 29.6 December-01 n.a n.a n.a n.a n.a n.a n.a n.a n.a


February-02 7.9 713.0 2.58 17.29 0.138 0.63 18.058 13.7 41.0 March-02 7.8 n.a 1.79 26.78 0.067 0.02 26.867 5.7 48.6

April-02 7.5 723.0 3.33 21.44 0.005 0.07 21.515 5.8 17.3 May-02 7.9 86.0 8.06 22.95 0.013 0.35 23.313 4.9 3.5

June-02 8.4 168.7 9.59 1.74 n.a 0.39 n.a 4.8 9.1 July-02 7.8 109.8 8.43 0.67 0.156 0.18 1.006 2.5 5.0




February-03 8.4 862.0 3.43 19.73 0.020 0.07 19.820 21.1 38.8 March-03 8.1 789.0 4.17 17.18 0.088 0.17 17.438 13.0 15.7

April-03 6.8 164.4 7.75 0.62 2.579 0.14 3.339 16.1 15.7 May-03 6.9 103.5 8.57 0.78 0.026 0.28 1.086 5.6 2.7

June-03 7.3 196.5 6.85 2.42 0.027 0.34 2.787 2.8 6.1 July-03 6.8 94.3 5.90 0.49 0.001 0.23 0.721 4.1 2.0


October-03 7.4 0.3 10.32 3.28 0.020 0.05 3.350 7.0 10.0 November-03 n.a n.a n.a n.a n.a n.a n.a n.a n.a December-03 8.0 0.98 18.16 0.017 0.08 18.257 n.a n.a


February-04 7.3 635.0 4.89 13.04 0.070 1.13 14.240 14.0 31.1 March-04 8.1 735.0 9.32 16.14 0.032 1.04 17.212 10.0 26.5

April-04 6.5 168.8 9.80 0.63 0.002 0.55 1.182 14.0 5.6 May-04 7.4 118.4 9.74 1.29 0.016 0.27 1.576 7.4 4.0

June-04 7.2 150.7 5.71 1.38 0.042 0.28 1.702 3.3 3.5 July-04 6.9 98.2 9.07 0.05 0.023 0.02 0.093 3.4 4.0


October-04 7.9 175.4 11.64 1.98 0.024 0.03 2.034 4.1 8.9 November-04 7.5 599.0 1.15 10.50 0.002 0.09 10.587 15.0 64< December-04 7.4 493.0 0.82 14.41 0.014 0.31 14.734 39.5 98.5


February-05 7.7 613.0 8.28 13.43 0.045 1.54 15.015 13.2 41.0 March-05 7.9 753.0 0.90 1.40 0.005 n.a n.a 37.0 62.8

April-05 7.2 712.0 2.61 12.62 0.325 0.24 13.185 39.0 37.4 May-05 7.9 141.9 8.39 1.64 0.074 0.32 2.034 5.0 2.4

June-05 7.2 134.1 9.28 1.04 0.050 0.15 1.240 6.6 6.6 July-05 7.2 161.7 8.48 0.94 0.045 1.11 2.095 4.4 6.6


October-05 7.9 236.0 10.16 5.94 0.048 0.35 6.338 5.2 5.1 November-05 7.9 645.0 4.00 14.32 0.048 0.54 14.908 29.6 31.6 December-05 7.7 649.0 2.40 13.33 0.094 0.56 13.984 20.4 51.6


February-06 7.5 773.0 n.a 21.76 0.045 1.69 23.495 n.a 55.4 March-06 8.2 721.0 0.77 20.29 0.339 0.09 20.719 38.4 44.2

April-06 7.3 695.0 0.46 7.12 0.084 0.19 7.394 30.4 43.8 May-06 6.9 94.1 8.62 0.37 0.010 0.44 0.820 12.0 2.3

June-06 6.9 104.7 7.85 0.92 0.028 0.16 1.108 6.0 4.8 July-06 7.6 120.5 7.55 0.53 0.065 0.36 0.955 4.5 1.3


October-06 7.0 406.0 4.44 8.32 0.122 8.442 34.7 20.6 November-06 7.6 789.0 5.98 12.16 0.165 0.87 13.195 14.5 20.6 December-06 7.7 876.0 4.70 21.78 0.085 0.63 22.495 27.8 39.7

Annual average -06 7.4 467.3 5.48 9.50 0.109 0.76 10.297 18.8 23.0 September-07 8.0 242.0 7.19 6.41 0.112 0.67 7.204 8.7 13.7

Hint: in last row, September 2007, pH, EC, DO measured during MSc field survey and samples for NH4

+, NO2-, NO3

-, COD-Mn, BOD5 collected by myself, analyzed in CLEM.



Appendix 5: Hydraulic dataset

Hydraulic measurement of the Tuul River, Ulaanbaatar station in 2006 Source: NAMHEM, Water sector

Water velocity by m s-1 Water depth by m N Date Direction

Water quality in direction

Water level by m (H)

Water discharge by m3 s-1 (Q)

Water area by m2

(F) mean max

River width by m mean max

Water surface gradient by %

Measured method

Analyzed method

1 21-Apr-06 1 very polluted 75 2.1 6.16 0.34 0.6 48.6 0.13 0.38 гр-21 4/4 задлаг

2 4-May-06 1 clean 131 25 30.3 0.82 1.37 64 0.47 0.8 гр-21 6/11 задлаг

3 12-May-06 1 clean 94 6.69 14.1 0.47 0.94 51.9 0.27 0.49 гр-21 5/7 задлаг

4 18-May-06 1 clean 142 41.2 39 1.06 1.7 66.5 0.59 0.92 гр-21 6/11 задлаг



7 1-Jun-06 1 clean 130 29 33.2 0.87 1.54 64.2 0.52 0.82 3.9 гр-21 6/10 задлаг

8 4-Jun-06 1 clean 178 69.9 64.7 1.08 1.38 buoy задлаг

9 6-Jun-06 1 clean 212 186.4 86.1 2.36 3.63 buoy задлаг

10 7-Jun-06 1 clean 197 124 78.9 1.57 2.1 buoy задлаг

11 9-Jun-06 1 clean 158 42 51 0.82 1.38 buoy задлаг

12 21-Jun-06 1 clean 124 21.5 28.3 0.76 1.33 61.9 0.46 0.75 4.1 гр-21 6/11 задлаг

13 24-Jun-06 1 clean 118 15.4 24.8 0.62 1.13 59.4 0.42 0.69 4.1 гр-21 6/10 задлаг

14 30-Jun-06 1 clean 146 45.2 41.1 1.1 1.78 67.2 0.61 0.95 4 гр-21 6/11 задлаг

15 1-Jul-06 1 clean 149 49.4 43 1.15 1.85 67.7 0.64 0.98 3.1 гр-21 6/11 задлаг

16 9-Jul-06 1 clean 172 53.8 57.8 0.93 1.25 buoy задлаг

17 10-Jul-06 1 clean 200 136 78.9 1.72 2.22 buoy задлаг

18 12-Jul-06 1 clean 190 106 71.8 1.48 1.9 buoy задлаг

19 17-Jul-06 1 clean 138 38.2 37.7 1.01 1.67 66 0.57 0.9 3 гр-21 6/11 задлаг



22 1-Aug-06 1 clean 147 48.2 42.6 1.13 1.79 67.6 0.63 0.98 4 гр-21 6/11 задлаг

23 3-Aug-06 1 clean 134 34 35.4 0.96 1.64 65.2 0.54 0.86 3.82 гр-21 6/11 задлаг



24 12-Aug-06 1 clean 120 17.4 26 0.67 1.15 60.1 0.43 0.71 3.72 гр-21 6/10 задлаг

25 15-Aug-06 1 clean 112 12.7 22.5 0.56 1.04 58.3 0.39 0.65 3.72 гр-21 6/9 задлаг

26 31-Aug-06 1 clean 107 10.2 20.4 0.5 0.92 57.4 0.36 0.63 3.63 гр-21 6/10 задлаг

27 4-Sep-06 1 clean 105 9.26 19.4 0.48 0.91 57.2 0.34 0.59 3.53 гр-21 6/10 задлаг



30 26-Sep-06 1 clean 110 11.6 21.9 0.53 1 58 0.38 0.64 3.63 гр-21 6/9 задлаг


32 2-Oct-06 1 clean 111 12.9 22.7 0.57 1.07 58.3 0.39 0.64 3.43 гр-21 6/9 задлаг




36 2-Nov-06 1 very polluted 96 6.11 14.6 0.42 0.82 52.5 0.28 0.52 3.43 гр-21 6/8 задлаг

37 10-Nov-06 1 very polluted 106 2.46 12 0.2 0.37 54.8 0.31 0.56 1.96 гр-21 5/5 задлаг





Appendix 6: CWTP chemical dataset

Month

Sam

plin

g po

int

Vol

ume,

tho

usan

d m

3

Wat

er

tem

pera

ture

, 0 C

pH

Tur

bidi

ty, c

m

Rem

aini

ng

capa

city

, ml

Susp

ende

d so

lid,

mg

l-1

BO

D5,

mg

l-1

BO

D20

, mg

l-1

DO

, mg

l-1

CO

D, m

g l-1

Alk

alin

ity,

g-e

cv l-1

Chl

orid

e, m

g l-1

NH

4+ , m

g l-1

NO

3- , mg

l-1

NO

2- , mg

l-1

PO

4-2, m

g l-1

Tot

al P

, mg

l-1

SO4-2

, mg

l-1

Tot

al C

r, m

g l-1

Tot

al N

, mg

l-1

EC

Col

or

incoming 14.00 8.76 2.20 3.17 289.00 301.50 450.00 859.03 4.20 71.02 30.18 4.76 1.25 3.13 84.79 1.30 28.30 614.00 5.07 January

discharging 4580

11.70 7.30 4.50 0.23 47.66 53.95 65.30 2.42 135.54 5.37 71.44 26.54 5.41 0.08 2.01 44.46 0.22 30.56 664.00 5.03

incoming 11.50 8.67 2.25 2.75 261.00 271.94 323.40 505.44 3.65 47.86 29.00 5.53 0.47 1.48 33.65 0.12 56.04 419.50 5.20 February

discharging 4219

9.00 7.25 11.25 33.00 38.60 53.23 4.82 112.32 4.26 43.4 16.79 7.81 0.08 1.14 25.08 0.02 16.81 407.50 4.55

incoming 10.00 8.49 2.26 2.93 461.61 257.74 360.00 737.71 4.80 103.13 30.36 1.35 0.33 1.40 58.21 0.64 55.48 615.33 5.17 March

discharging 4702

8.67 7.01 10.40 35.33 30.60 42.52 4.41 139.10 4.07 60.52 29.33 3.21 0.01 0.81 35.43 0.16 30.26 502.33 4.63

incoming 14.00 7.45 2.43 2.33 377.33 243.21 329.00 603.72 3.89 79.84 33.59 1.86 0.15 2.16 30.95 0.54 23.82 608.00 5.73 April

discharging 4896

13.00 7.01 6.30 39.66 24.13 37.00 2.3 140.51 3.55 61.29 18.72 3.66 0.01 1.40 22.09 0.35 14.29 523.67 4.90

incoming 13.00 7.47 3.00 1.53 257.30 257.10 430.00 332.27 3.95 71.03 19.93 3.22 0.03 0.81 1.50 0.77 534.30 5.70 May

discharging 4532

12.00 7.13 15.90 18.00 17.89 41.40 3.54 83.03 3.53 57.18 16.50 4.34 0.01 0.58 1.33 0.06 486.70 4.50

incoming 17.00 6.84 2.55 2.25 247.50 300.50 375.00 582.88 3.55 62.88 23.82 3.68 0.00 0.30 38.80 0.31 547.00 4.60 June

discharging 4233

15.00 6.85 10.20 15.50 33.20 40.80 1.99 111.02 3.63 46.93 15.74 4.94 0.00 0.13 27.84 0.00 517.50 4.10

incoming 19.00 6.62 2.75 3.25 297.67 325.65 496.99 3.97 72.58 20.08 7.07 0.06 1.55 1.15 37.33 0.18 600.00 5.17 July

discharging 4330

17.00 6.71 17.25 38.00 30.99 1.87 85.70 3.39 53.56 21.50 8.00 0.00 1.52 0.63 24.80 0.16 558.00 4.67

incoming 20.00 6.92 2.00 3.96 405.00 284.22 530.14 4.41 96.08 29.69 4.49 0.36 2.51 38.03 0.14 687.00 5.30 August

discharging 4550

18.00 6.83 7.40 25.30 29.12 1.80 86.98 3.96 71.13 21.30 5.41 0.05 1.92 21.48 0.00 685.00 5.50

incoming 18.00 7.68 2.73 3.20 288.00 233.93 457.68 4.96 109.75 24.25 2.31 0.10 2.23 64.00 0.44 667.50 5.40 September

discharging 4790

16.00 6.98 7.90 63.60 33.80 1.39 95.55 4.24 92.04 21.01 3.63 0.09 1.31 50.73 0.17 703.70 7.43

incoming 18.00 7.95 3.07 3.03 529.33 264.77 552.96 5.01 101.68 27.08 2.68 1.09 2.86 68.28 0.21 645.00 5.73 October

discharging 5188

16.00 6.70 11.37 67.50 32.14 118.59 4.05 78.03 16.90 5.08 0.03 1.12 32.42 0.05 638.67 5.07

incoming 14.00 8.10 2.80 3.50 406.67 297.27 482.98 4.43 92.17 24.17 3.64 1.11 2.68 3.41 0.48 681.33 8.23 November

discharging 4849

13.00 6.92 11.00 45.33 36.03 6.93 78.98 3.94 73.53 19.37 4.54 0.05 1.00 2.27 0.12 612.33 5.45

incoming 14.00 8.67 2.60 3.33 468.00 345.35 267.36 5.39 83.84 23.17 2.83 1.57 3.13 3.00 0.93 595.33 6.20 December

discharging 5215

11.30 7.13 7.50 48.00 43.35 126.72 4.15 67.99 15.55 5.33 0.03 1.70 2.41 0.09 621.33 5.17

incoming 15.21 7.80 2.55 2.94 357.37 281.93 534.10 4.35 82.66 26.28 3.62 0.54 2.02 2.27 50.45 0.51 40.91 601.19 5.63 Mean

discharging 4674

13.39 6.99 10.08 0.02 39.74 33.65 3.15 109.50 4.01 64.75 19.94 5.11 0.04 1.22 1.66 31.59 0.12 22.98 576.73 5.08



Appendix 7: A flow model setup

Upper CS – Zaisan Trapezoid Floor width=48.6 Slope=5.71 Max depth=1.5 Floor level=285.25 Surface level=292.6 Sonsgolon CS Trapezoid Floor width=44.5 Slope=5.65 Max depth=1.5 Floor level=265.44 Surface level=272.12 Songino (upper) CS Trapezoid Floor width=41.3 Slope=5.61 Max depth=1.5 Floor level=249.6 Surface level=255.73 Songino (down) CS Trapezoid Floor width=41.0 Slope=5.6 Max depth=1.5 Floor level=247.62 Surface level=253.69 Chicken farm CS Trapezoid Floor width=36.8 Slope=5.54 Max depth=1.5 Floor level=227.81 Surface level=233.2 Khadanhyasaa CS Trapezoid Floor width=33.5 Slope=5.5 Max depth=1.5 Floor level=211.97 Surface level=216.82 Altanbulag CS Trapezoid Floor width=26.6 Slope=5.4 Max depth=1.5 Floor level=178.3 Surface level=182 Lower CS – Extension Trapezoid Floor width=10 Slope=20 Max depth=1.5 Floor level=173.3 Surface level=177

BC upstream – Zaisan Q Starting

05.3.1-06.12.1 Q=0.033 H = 286.44=285.25+1.19 Location: 643389 E, 5305700 N

BC midstream - none Calibration point using Q at Altanbulag ≈ 54 km far from upstream

BC downstream – Artificial Q Starting

05.3.1-06.12.1 Q=-1.38

10 days constant H = 174.8=173.3+1.5

Location: 587593 E, 5301654 N 20 km far from Altanbulag

Section - Tuul ≈ 54 km long CS interpolated Min distance 10 m and max 200

m n=0.06

DP – CWTP Q Starting

05.3.1-06.12.1 Q=1.814 Location: 17.5 km far from upstream

DP – Bio Q Starting

05.3.1-06.12.1 Q=0.0172 Location: 2.5 km far from CWTP

Initial condition Q 0.198 1.38 H 286.44 174.8

Calculation Starting

05.3.1-06.12.1 Time step=1” Theta=0.9 Distance=200 Total

Manning Summer: 4.1 – 12.1



Appendix 8: Manning value

Manning's n for Channels (Chow, 1959)

Type of Channel and Description Minimum Normal Maximum

Natural streams - minor streams (top width at flood stage < 100 ft)

1. Main Channels

a. clean, straight, full stage, no rifts or deep pools 0.025 0.030 0.033

b. same as above, but more stones and weeds 0.030 0.035 0.040

c. clean, winding, some pools and shoals 0.033 0.040 0.045

d. same as above, but some weeds and stones 0.035 0.045 0.050

e. same as above, lower stages, more ineffective slopes

and sections 0.040 0.048 0.055

f. same as "d" with more stones 0.045 0.050 0.060

g. sluggish reaches, weedy, deep pools 0.050 0.070 0.080

h. very weedy reaches, deep pools, or floodway

with heavy stand of timber and underbrush 0.075 0.100 0.150

2. Mountain streams, no vegetation in channel, banks usually steep, trees and brush along banks

submerged at high stages

a. bottom: gravels, cobbles, and few boulders 0.030 0.040 0.050

b. bottom: cobbles with large boulders 0.040 0.050 0.070

3. Floodplains

a. Pasture, no brush

1.short grass 0.025 0.030 0.035

2. high grass 0.030 0.035 0.050

b. Cultivated areas

1. no crop 0.020 0.030 0.040

2. mature row crops 0.025 0.035 0.045

3. mature field crops 0.030 0.040 0.050

c. Brush

1. scattered brush, heavy weeds 0.035 0.050 0.070

2. light brush and trees, in winter 0.035 0.050 0.060

3. light brush and trees, in summer 0.040 0.060 0.080

4. medium to dense brush, in winter 0.045 0.070 0.110

5. medium to dense brush, in summer 0.070 0.100 0.160

d. Trees

1. dense willows, summer, straight 0.110 0.150 0.200

2. cleared land with tree stumps, no sprouts 0.030 0.040 0.050

3. same as above, but with heavy growth of sprouts 0.050 0.060 0.080

4. heavy stand of timber, a few down trees, little

undergrowth, flood stage below branches 0.080 0.100 0.120

5. same as 4. with flood stage reaching branches 0.100 0.120 0.160

Source: http://www.fsl.orst.edu/geowater/



Appendix 9: Quality model syntax

(Source: Department of Water Quality Management, University of Wageningen)

/* River water quality model for the DO */ /* In case of river Tuul, Ulaanbaatar city, Mongolia */ /* MSc research */ /* WREM, ITC, 2008 */ WATER DO [8.0] mg/l ; Dissolved oxygen WATER BOD [5.0] mg/l ; Biochemical oxygen demand WATER NH4 [0.05] mg/l ; Ammonium PARM Beta [0.0001] (gO2/mg Chl)/ (W/m2) ; Oxygen production constant (photosynthesis) PARM fd [1.0] - ; Fraction of dissolved BOD PARM Kd [0.1] day ; BOD degradation rate constant PARM KDO [1.0] mgO2/l ; Monod constant DO inhibition BOD decay PARM KLmin [0.1] m/day ; Minimum oxygen mass-transfer rate in flowing system PARM KNDO [1.0] mgO2/l ; Monod constant DO inhibition nitrification PARM Knit [1.0] day ; Nitrification rate constant PARM OPTKL [1.0] - ; Option material transfer (0=stagnation 1=flowing) PARM TKd [1.05] - ; Temperature coefficient of BOD decay PARM TKL [1.024] - ; Temperature coefficient of oxygen mass-transfer PARM TKnit [1.05] - ; Temperature coefficient of nitrification PARM TSOD [1.06] - ; Temperature coefficient of sediment oxygen demand PARM Vs [1.0] m/day ; Settling velocity of BOD XT T [15.0] oC ; Water temperature XT SOD [2.0] g/m2, day ; Sediment oxygen demand XT SNH4 [0.0] g/m2, day ; Diffusion of NH4 concentration XT SBOD [0.0] g/m2, day ; Diffusion of BOD concentration XT I0 [0.00] W/m2 ; Solar radiation XT A [10.0] ug/l ; Algal biomass XT W [3.0] m/s ; Wind speed FLOW Q [10.0] m3/s ; Discharge FLOW As [60.0] m2 ; Cross-section of flow area FLOW Z [1.0] m ; Water depth U=ABS(Q/As); OS=14.652-0.41022*T+0.007991*T*T-0.000077774*T*T*T; IF (OPTKl==0)

{Kl20=0.0864*(8.43*W^0.5-3.67*W+0.43*W*W);} IF (W<1.82) {Kl20=0.37+0.09*W;} IF (OPTKl==1) {Kl20=2.33*U^0.67*Z^(-0.85);} IF (Kl20 <Klmin) {Kl20=Klmin;} KlT=Kl20*TKl^(T-20); KA=KlT/Z; BODU=BOD/(1-EXP(-Kd*5)); PDO=Beta*I0*A; SedDO=-SOD*TSOD^(T-20)/Z; REAR=KA*(OS-DO); Nitrif=-4.57*Knit*NH4*TKnit^(T-20)*DO/(DO+KNDO); BODox=-Kd*BODU*TKd^(T-20)*DO/(DO+KDO); k1(DO)=-KA; k0(DO)=KA*OS+PDO+SedDO+BODox+Nitrif; k1(BOD)=-Vs*(1-fd)/Z-Kd*TKd^(T-20)*DO/(DO+KDO); k0(BOD)=SBOD/Z; k1(NH4)=-Knit*TKnit^(T-20)*DO/(DO+KNDO); k0(NH4)=SNH4/Z;



Appendix 10: A quality model setup

Upper CS – Zaisan Trapezoid Floor width=48.6 Slope=5.71 Max depth=1.5 Floor level=285.25 Surface level=292.6

Sonsgolon CS Trapezoid Floor width=44.5 Slope=5.65 Max depth=1.5 Floor level=265.44 Surface level=272.12

Songino (upper) CS Trapezoid Floor width=41.3 Slope=5.61 Max depth=1.5 Floor level=249.6 Surface level=255.73

Songino (down) CS Trapezoid Floor width=41.0 Slope=5.6 Max depth=1.5 Floor level=247.62 Surface level=253.69

Chicken farm CS Trapezoid Floor width=36.8 Slope=5.54 Max depth=1.5 Floor level=227.81 Surface level=233.2

Khadanhyasaa CS Trapezoid Floor width=33.5 Slope=5.5 Max depth=1.5 Floor level=211.97 Surface level=216.82

Altanbulag CS Trapezoid Floor width=26.6 Slope=5.4 Max depth=1.5 Floor level=178.3 Surface level=182

Lower CS – Extend Trapezoid Floor width=10 Slope=20 Max depth=1.5 Floor level=173.3 Surface level=177

BC upstream – Zaisan Q=0.033 BOD=6.0 DO=9.94 NH4-=0.7 Location: 643389 E, 5305700 N

BC - Sonsgolon Q=0.172 BOD=6.1 DO=8.48 NH4-=0.78 Location: 633139 E, 5303905 N

BC – Songino (upper) Q=0.155 BOD=6.0 DO=7.42 NH4-=1.51 Location: 626450 E, 5301560 N

BC – Songino (down) calibration Location: 625306 E, 5300735 N

BC – Chicken farm Q=2.025 BOD=26.9 DO=1.14 NH4-=1.94 Location: 619858 E, 5292251 N

BC – Khadanhyasaa Q=2.0 BOD=6.9 DO=4.4 NH4-=1.58 Location: 612478 E, 5289880 N

BC – Altanbulag Q=0.01 BOD=6.0 DO=7.5 NH4-=1.05 Location: 597134 E, 5283564 N

BC downstream – Artificial Q=-1.38 BOD=6.0 DO=7.5

NH4-=1.05 Location: 587593 E, 5301654 N

20 km far from Altanbulag

Section - Tuul 54 km long CS interpolated Min distance 10 m and max 200 m n=0.06

DP – CWTP Q=1.814 BOD=30.6 DO=4.41 NH4=29.33

DP – Bio Q=0.0172 BOD=30.6 DO=4.41 NH4=29.33

Calculation Starting

05.3.1-06.12.1 Time step=30’ Theta=0.55 Distance=200 Output = DO



Appendix 11: Required input data for the quality model

Type Abbreviation Source Typical value

DO CLEM

BOD5 CLEM Initial condition

NH4+ CLEM

DO CLEM

BOD5 CLEM Boundary condition

(system) NH4

+ CLEM

DO CLEM

BOD5 CLEM Boundary condition

(middle nodes) NH4

+ CLEM

KLmin Literature 0.1 m.day-1

TKL Literature 1.024

Kd Literature 0.1-0.3 day

Vs Literature < 1 m.day-1

fd Literature 0.8-1.0

KDO Literature 1 mg.l-1

TKd Literature 1.03-1.05

Knit Literature 0.1-1.0 day

TKnit Literature 1.05-1.10

KNDO Literature 1.0-2.0 mg.l-1

Beta Literature 0.0001-0.001

g day-1.mgChl-1.(W.m-2)-1

TSOD Literature 1.06

Parameters

OPTKL Literature 0=stagnant, 1=running

D Literature/estimate

T CLEM

SBOD Estimate

SNH4 Estimate

IO IWEC

A Estimate 10-200 ug.l-1

SOD Literature 0.5-2.0 g.m-2.day-1

External variable

W IWEC

Source: [Lijklema, et al., 1996]